Intro to Electric Powertrains
04-12-2017, 08:15 AM
#1
Instructor
Thread Starter
Intro to Electric Powertrains
Hi all!
As a Porsche enthusiast, I usually lurk on the 996 forums, though I didn't sign up for an account until I needed some headlight info for a personal project, but in my day job, I'm a battery system engineer at an electric vehicle startup that's gaining a bit of traction (no pun intended).
There's a lot of misinformation out there about electric cars and how and why they work, so before we get too bogged down in rumors about the Mission E (which I'm very excited about - ya'll ain't seen nothing yet), I'd like to clear things up a bit and provide some useful background.
Please note that this is a very long post, and is a summary of what I've been working on the last several years of my life. I don't go into all of the math or specific concepts behind everything, but I do try to give a general idea of what to keep in mind when you're hearing all of the electric vehicle hype over the next few years.
TL;DR: the main points that I go into are
The number one thing you need to know about electric powertrains is that nobody fully understands the electric powertrain.
Starting at about 200 HP at the motor, and 40kWh pack capacity, everybody is pretty much doing their own thing. Below that, it's pretty reasonable to expect an off-the-shelf solution to match your performance needs while fitting within the bounds of a passenger car.
When you're trying to push 400+ HP in a motor that's only slightly larger than a basketball, you have two options for motor technology: permanent magnet, and induction. Both of these require 3-phase AC power. You only know which is better for your purpose when you've done Watson-months of analysis of your entire system, from the chemistry of the battery to the roads you select for your use case.
Yes, I said 3-phase AC power. Out of a battery which supplies DC. You need an inverter, too, and said inverter needs to be able to deliver as much power as you need momentarily - the peak power rating, typically component-limited - and over some defined amount of time - the RMS or "mean" power rating, which is typically cooling-limited. At 400HP peak, you're looking at a ream of paper in size at best - and I want to be very clear here, what I mean by "at best" is a sufficiently sized team of the best engineers in the world with an effectively unlimited budget for several years.
Oh, and your battery, switches, busbars, cables, etc. also have peak and RMS power limits. More on that later.
The number two thing you need to know about the electric powertrain is that there is a linear relationship between power and range.
If you have a 100kWh+ battery pack, you will be able to deliver 1000HP, and you will have a 300+ mile range, period. At that pack size, those both come for free.
Here's why.
Power is voltage times current. Voltage scales with the number of cells you have in series ("S count"). This should sound familiar from basic electronics. If you pull apart a 9V battery, you'll see 6 AAAA cells connected positive end to negative end in a line - in series. The AAAA cell, like a AAA, AA, A, B, C, and D cell, delivers current at around 1.5V.
The inside of a battery is a soup undergoing two-ish chemical reactions, and these reactions are limited by the electrons that can move from one end to the other. The electrons are prevented from going directly through this soup (mostly, and the exceptions are complicated), and instead go around through a conductive path. This conductive path, and the chemical reaction itself, both have some amount of resistance.
The main losses in any electric system are the direct conversion of electricity to heat. In the industry, we call these "resistive losses" or "I squared R losses" - the formula is P = I^2 * R. This heat either goes somewhere, or the temperature of the resistive element increases. As an aside, if you only draw a ton of current for a moment - let's say we draw the most the system can deliver, and call this "peak" current, you don't actually heat up and my job gets really boring really quickly.
Now, I mentioned that there is some resistance in the battery itself. There is a chemistry-dependent temperature at which BAD THINGS HAPPEN to lithium-ion chemistries. We call this "thermal runaway", and it leads to things like
. Effectively, this means that we can't let the batteries get too hot, so we need to cool them and limit our resistive losses. Since resistance is (more or less) constant, this means that we need to limit the amount of current each cell sees.
So, to sum up these last few paragraphs, your system voltage scales linearly with how many cells you have in series, and your system's maximum current scales linearly with how many cells you have in parallel, so your system's power scales linearly with how many cells you have.
Now, power is also the force your system is applying times the momentary speed. Assuming we are in a steady state situation at constant highway speed - which makes the math easier to follow, but trust me it works in dynamic systems as well - the main force acting on your car is air resistance, which mostly varies with the square of speed. In this case, the amount of energy you use in a given time is your steady-state power times the time duration in question. This just so happens to equal force times distance - so in a steady-state situation, the [distance you travel] equals your [range] equals [power times time].
This also equals [(voltage times current) times time], or re-emphasized, [voltage times (current times time)].
So back to batteries. If you look at a generic 18650 lithium cell, you might see an advertised "capacity" of 3000mAh - in theory, the cell can deliver 3000mA for a duration of 1 hour. (Note: 9/10 times, this cannot be achieved. Yes, they're lying to you. No, you can't do anything about it.) Two of these cells in parallel would be able to deliver 6000mA for 1 hour.
Yes, this means that your range scales with your voltage times your capacity - in other words, it scales with the number of cells you have in series times the number of cells you have in parallel. Just like power does.
The number three thing you need to know about electric powertrains is that motor power and battery pack power are different.
For the sake of significantly easier math, we're assuming a perfect DC brushed motor and steady state systems for this entire section.
The torque an electric motor delivers is the current flowing through the windings (with some losses from the resistance in the windings) times a constant. This constant, reasonably, is defined in units of Torque/Current.
The voltage drop across this motor is the speed the motor is spinning (plus a squidge, again because of the resistance in the windings) times a constant. This constant, also reasonably, is defined in units of Voltage/Angular Speed.
Please also note that power, in one of its many iterations, is Torque times Angular Speed. Therefore:
Torque / Current * Angular Speed / Angular Speed
= Torque * Angular Speed / (Current * Angular Speed)
= Power / (Current * Angular Speed) = Power/Current / Angular Speed
= Voltage / Angular Speed
Depending on what units you pick, these two constants can be equal. SI is one of the systems that allows this - 1Nm/A = 1V/(rad/s) (note that the radian is dimensionless).
The power the motor takes from the system is the voltage drop across the motor times the current flowing through the motor, and the motor delivers a decent fraction of this power to the transmission (90-95% dependent on a lot of factors).
Now, remember how I said batteries deliver current at some voltage? Let's look at the three basic laws in electronics:
When the motor is not moving, the current flowing through the motor is the voltage of the source divided by the resistance of the resistor, in this case 12A. You cannot deliver more current without either raising the voltage or lowering the resistance. Remember, torque is proportional to current, so you have also reached the peak torque of the system - 0.12Nm - but as the motor isn't able to move, the power delivered by the motor is zero! However, the resistor is dissipating 12A * 12V = 144W as heat, and the voltage source is delivering 12V * 12A = 144W - this is the peak power the source can deliver in this system.
Let's have the motor spin as fast as it can. The highest voltage possible across the motor is the voltage of the source - due to Kirchhoff's Second Law - so we're looking at 12V / (0.001 V/(rad/s)) = 1200 rad/s, or 11460RPM. Quick test for this case: how much torque can we deliver at this speed?
The answer is none - we cannot flow any current, because we cannot have any voltage drop across the resistor, and 0A * 0.01Nm/A = 0Nm. This means that the power delivered by the motor in this case is also zero! In this case, though, the resistor is not dissipating any power as heat, and the voltage source is not delivering any power.
Let's now look at some point in the middle - let's say, 6A through the loop. The resistor is dissipating 6A * 6A * 1Ohm = 36W, and has a voltage drop of 6A * 1Ohm = 6V. The motor is delivering 0.01Nm/A * 6A = 0.06Nm, and has a voltage drop of 12V - 6V = 6V, so is spinning at 6V/(0.01V/(rad/s)) = 600 rad/s. As such, the motor is delivering a power of 6A * 6V = 36W. Just to double-check our numbers, 0.06Nm * 600 rad/s = 36W - we do not need to know both what the motor is doing and what its voltage and current are. The source is delivering 36W + 36W = 72W.
As a final exercise in this idealized system, let's determine the peak power the motor can deliver, assuming steady state conditions. Fortunately, we learn a couple interesting things on the way. Unfortunately, we need to do some calculus.
P_motor = V_motor * I_system
= (V_source - V_res) * I_system
= V_source * I_system - V_res * I_system
= V_source * I_system - I_system^2 * R_res
The power delivered by the motor equals the power delivered by the source, less resistive losses. This makes sense - power has to come from and go somewhere in a steady state system - but note for later that it doesn't matter where in the system the resistive losses are.
Continuing, the peak power is where either slightly more current or slightly less delivers lower power - the slope of the power-current curve is positive towards lower current, and negative towards more current, so we want the slope to equal zero, or in other words we need a point such that the derivative of power with respect to current equals zero.
dP_motor / dI_system = V_source - 2 * I_system * R_res = 0
V_source = 2 * I_system_motpeak * R_res
I_system_motpeak = V_source / R_res / 2 = 12V / 1Ohm / 2 = 6A
P_motor_peak = 12V * 6A - 6A * 6A * 1Ohm = 72W - 36W = 36W
It's like I planned this or something. Another important takeaway: the voltage source does not care whether the motor is a motor delivering some amount of torque at some speed, or a black box with a resistor appropriately sized to have some voltage drop at some current.
Now let's look at this in conjunction with a battery, which conveniently can be modeled as a voltage source in series with a resistor.
The Tesla Model S P90D has a 96S74P battery pack rated to 90kWh capacity. The cells can be charged to 4.3V, the internal resistance is on the order of 0.035Ohm, the wire bonds connecting the cells to the busbars in the modules can tolerate about 35A each without breaking, and where applicable let's assume the motor and inverter are each 95% efficient. The wire bonds themselves (1 on each end of each cell) are about 0.007Ohm. All busbar, pack fuse, connector, etc. resistances are assumed to be zero (and are in fact pretty close to negligible).
V_max = 4.3V * 96 = 412.8V
I_max = 35A * 74 = 2590A (note: this would blow their pack-level fuses in well under a second)
R_cells = 0.035Ohm / 74 * 96 = 0.045Ohm
R_pack = 0.049Ohm / 74 * 96 = 0.064Ohm
The peak power the chemical reactions in the cells are allowed to deliver is 412.8V * 2590A = 1070kW = 1434HP. The resistive losses within the cells are 2590A ^2 * 0.045Ohm = 302kW (405HP), and the resistive losses within the pack as a whole are 2590A ^2 * 0.064Ohm = 430kW.
As such, the pack can deliver a peak power of 640kW = 860HP.
It is extremely important to note that there is still a lot of current headroom based on the voltage and resistance of the pack - if you short across the pack terminals with closed contactors, you will momentarily see about
412.8V / 0.064Ohm = 6450A
before the wire bonds and pack-level fuse all blow and no more current flows. This means that we can deliver maximum current until the voltage drop across the inverter equals
412.8 V - 2590A * 0.064Ohm = 247 V
If we ask for the maximum current given an arbitrary voltage across the inverter terminals (assuming the only resistance in the system is in the pack):
I_inverter = I_system = (V_max - V_inverter) / R_pack
P_inverter = I_inverter * V_inverter
= V_max * V_inverter / R_pack - V_inverter ^2 / R_pack
If we then solve for the maximum power the inverter can take from the system (and for the sake of math I can actually demonstrate in a forum post, let's assume that corresponds to both peak inverter power delivery and peak motor power delivery), we take the derivative of inverter power with respect to inverter voltage:
dP_inverter/dV_inverter = V_max / R_pack - 2 * V_inverter / R_pack = 0
V_max / R_pack = 2 * V_inverter / R_pack
V_inverter = V_max / 2 = 412.8 V /2 = 206.4 V
Note that this is a lower voltage than the boundary condition imposed by allowable motor current! We therefore have two options for our peak power: 206.4V * 2590A or 247V * 2590A. The latter is significantly higher value, at 640kW or 858HP.
Side note: in this particular case, we can maintain our peak system torque more than halfway through the motor's speed profile. It's not quite that impressive in the real world, but there is a range of speeds at which a real motor can deliver peak allowable torque, and at max allowable speed there is a range of deliverable torques. These ranges are defined by thermal and mechanical limits respectively, and the motor constant really doesn't come into play.
Now, inverter and motor efficiencies both being 95% means the motor actually delivers a peak power of 858HP * 0.95 * 0.95 = 774HP. Tesla claims 762HP for the P90D, which could easily be accounted for by some resistance in busbars, connectors, etc., my assumed values being slightly off, and so forth.
However, this does not account for drivetrain and tire rolling efficiency - which we can reasonably take as 90% and 97% efficient respectively, for a power at the road of 676HP.
The number four thing you need to keep in mind about the electric powertrain is that battery properties are not constant.
Open-circuit (no-current) voltage drops as the state of charge decreases. Real-world capacity decreases as the rate of current draw increases. Capacity drops and internal resistance rises as cells are subjected to charge/discharge cycles, and these changes occur significantly faster during higher-current and low-temperature charge and discharge cycles.
What this means for the average consumer is that supercharging is not something you want to do all the time, you should treat your electric vehicle with care, and understand that your battery will likely be a significant influence on long-term resale value and maintenance costs.
The number five thing to keep in mind about the electric powertrain is that speed is problematic.
Back in section 2, I mentioned that air resistance is proportional to the square of speed, and at highway speeds is the main force acting against a vehicle's motion. Remember that power is force times speed - which means that, for all intents and purposes, the power a vehicle requires to maintain speed is proportional to the cube of speed - this means that your engine needs to put out 8 times as much power at 150mph as at 75mph.
Let's take an electric vehicle that has a 90kWh battery pack and takes 10kW to go 60mph. It follows that said vehicle can go 540 miles on a charge! As long as it doesn't speed up, slow down, go up a hill, run the air conditioner, run the heater, etc.
If instead the driver wanted to go the speed of traffic of 75mph, the required power is 19.5kW and the range drops to (75mph * 90kWh / 19.5kW) = 346mi.
Now let's say we're on the Autobahn. We're going 155mph in this car, requiring 172.4kW for a range of... 81 miles. If this is the mechanical top speed in a gear, a peak motor power of around 560kW is delivered around 40mph.
One more case: we pick a gear ratio that gives us peak power of 760HP=567kW at top speed. Our top speed is then 230.5mph, and we can go 37 miles. (Note that, because peak power drops as state of charge drops, we won't be able to reach that speed, nor can we maintain any top speed).
That particular case isn't very realistic, though. With a one-gear system, and assuming peak power is at about 95% of peak current, peak current would be maintained from 0-200mph - which has severe thermal implications - and the motor would be delivering about 50HP to accelerate at 40mph, thus making a Geo Metro seem sprightly.
So, why not use a multi-gear system? Efficiency. Specifically, a shifting mechanism costs about 5% drivetrain efficiency, which loosely translates to 15 miles in a 300-mile-range car - and range anxiety is one of the primary factors slowing the growth of the electric car market. The average passenger car stays below 80mph for upwards of 95% of its mileage, and because electric cars are incredibly fast in this range, manufacturers emphasize that instead (while ensuring that the Autobahn can be driven at appropriate speeds).
As a Porsche enthusiast, I usually lurk on the 996 forums, though I didn't sign up for an account until I needed some headlight info for a personal project, but in my day job, I'm a battery system engineer at an electric vehicle startup that's gaining a bit of traction (no pun intended).
There's a lot of misinformation out there about electric cars and how and why they work, so before we get too bogged down in rumors about the Mission E (which I'm very excited about - ya'll ain't seen nothing yet), I'd like to clear things up a bit and provide some useful background.
Please note that this is a very long post, and is a summary of what I've been working on the last several years of my life. I don't go into all of the math or specific concepts behind everything, but I do try to give a general idea of what to keep in mind when you're hearing all of the electric vehicle hype over the next few years.
TL;DR: the main points that I go into are
- Nobody fully understands the electric powertrain.
- There is a linear relationship between power and range.
- Motor power and battery pack power are different.
- Battery properties are not constant.
- Speed is problematic for electric vehicles.
The number one thing you need to know about electric powertrains is that nobody fully understands the electric powertrain.
Starting at about 200 HP at the motor, and 40kWh pack capacity, everybody is pretty much doing their own thing. Below that, it's pretty reasonable to expect an off-the-shelf solution to match your performance needs while fitting within the bounds of a passenger car.
When you're trying to push 400+ HP in a motor that's only slightly larger than a basketball, you have two options for motor technology: permanent magnet, and induction. Both of these require 3-phase AC power. You only know which is better for your purpose when you've done Watson-months of analysis of your entire system, from the chemistry of the battery to the roads you select for your use case.
Yes, I said 3-phase AC power. Out of a battery which supplies DC. You need an inverter, too, and said inverter needs to be able to deliver as much power as you need momentarily - the peak power rating, typically component-limited - and over some defined amount of time - the RMS or "mean" power rating, which is typically cooling-limited. At 400HP peak, you're looking at a ream of paper in size at best - and I want to be very clear here, what I mean by "at best" is a sufficiently sized team of the best engineers in the world with an effectively unlimited budget for several years.
Oh, and your battery, switches, busbars, cables, etc. also have peak and RMS power limits. More on that later.
The number two thing you need to know about the electric powertrain is that there is a linear relationship between power and range.
If you have a 100kWh+ battery pack, you will be able to deliver 1000HP, and you will have a 300+ mile range, period. At that pack size, those both come for free.
Here's why.
Power is voltage times current. Voltage scales with the number of cells you have in series ("S count"). This should sound familiar from basic electronics. If you pull apart a 9V battery, you'll see 6 AAAA cells connected positive end to negative end in a line - in series. The AAAA cell, like a AAA, AA, A, B, C, and D cell, delivers current at around 1.5V.
The inside of a battery is a soup undergoing two-ish chemical reactions, and these reactions are limited by the electrons that can move from one end to the other. The electrons are prevented from going directly through this soup (mostly, and the exceptions are complicated), and instead go around through a conductive path. This conductive path, and the chemical reaction itself, both have some amount of resistance.
The main losses in any electric system are the direct conversion of electricity to heat. In the industry, we call these "resistive losses" or "I squared R losses" - the formula is P = I^2 * R. This heat either goes somewhere, or the temperature of the resistive element increases. As an aside, if you only draw a ton of current for a moment - let's say we draw the most the system can deliver, and call this "peak" current, you don't actually heat up and my job gets really boring really quickly.
Now, I mentioned that there is some resistance in the battery itself. There is a chemistry-dependent temperature at which BAD THINGS HAPPEN to lithium-ion chemistries. We call this "thermal runaway", and it leads to things like
So, to sum up these last few paragraphs, your system voltage scales linearly with how many cells you have in series, and your system's maximum current scales linearly with how many cells you have in parallel, so your system's power scales linearly with how many cells you have.
Now, power is also the force your system is applying times the momentary speed. Assuming we are in a steady state situation at constant highway speed - which makes the math easier to follow, but trust me it works in dynamic systems as well - the main force acting on your car is air resistance, which mostly varies with the square of speed. In this case, the amount of energy you use in a given time is your steady-state power times the time duration in question. This just so happens to equal force times distance - so in a steady-state situation, the [distance you travel] equals your [range] equals [power times time].
This also equals [(voltage times current) times time], or re-emphasized, [voltage times (current times time)].
So back to batteries. If you look at a generic 18650 lithium cell, you might see an advertised "capacity" of 3000mAh - in theory, the cell can deliver 3000mA for a duration of 1 hour. (Note: 9/10 times, this cannot be achieved. Yes, they're lying to you. No, you can't do anything about it.) Two of these cells in parallel would be able to deliver 6000mA for 1 hour.
Yes, this means that your range scales with your voltage times your capacity - in other words, it scales with the number of cells you have in series times the number of cells you have in parallel. Just like power does.
The number three thing you need to know about electric powertrains is that motor power and battery pack power are different.
For the sake of significantly easier math, we're assuming a perfect DC brushed motor and steady state systems for this entire section.
The torque an electric motor delivers is the current flowing through the windings (with some losses from the resistance in the windings) times a constant. This constant, reasonably, is defined in units of Torque/Current.
The voltage drop across this motor is the speed the motor is spinning (plus a squidge, again because of the resistance in the windings) times a constant. This constant, also reasonably, is defined in units of Voltage/Angular Speed.
Please also note that power, in one of its many iterations, is Torque times Angular Speed. Therefore:
Torque / Current * Angular Speed / Angular Speed
= Torque * Angular Speed / (Current * Angular Speed)
= Power / (Current * Angular Speed) = Power/Current / Angular Speed
= Voltage / Angular Speed
Depending on what units you pick, these two constants can be equal. SI is one of the systems that allows this - 1Nm/A = 1V/(rad/s) (note that the radian is dimensionless).
The power the motor takes from the system is the voltage drop across the motor times the current flowing through the motor, and the motor delivers a decent fraction of this power to the transmission (90-95% dependent on a lot of factors).
Now, remember how I said batteries deliver current at some voltage? Let's look at the three basic laws in electronics:
- Ohm's Law: Voltage equals current times resistance
- Kirchhoff's First Law: The amount of current flowing into a point/node/junction equals the amount of current flowing out
- Kirchhoff's Second Law: The sum of the voltage changes around any closed network is zero
When the motor is not moving, the current flowing through the motor is the voltage of the source divided by the resistance of the resistor, in this case 12A. You cannot deliver more current without either raising the voltage or lowering the resistance. Remember, torque is proportional to current, so you have also reached the peak torque of the system - 0.12Nm - but as the motor isn't able to move, the power delivered by the motor is zero! However, the resistor is dissipating 12A * 12V = 144W as heat, and the voltage source is delivering 12V * 12A = 144W - this is the peak power the source can deliver in this system.
Let's have the motor spin as fast as it can. The highest voltage possible across the motor is the voltage of the source - due to Kirchhoff's Second Law - so we're looking at 12V / (0.001 V/(rad/s)) = 1200 rad/s, or 11460RPM. Quick test for this case: how much torque can we deliver at this speed?
The answer is none - we cannot flow any current, because we cannot have any voltage drop across the resistor, and 0A * 0.01Nm/A = 0Nm. This means that the power delivered by the motor in this case is also zero! In this case, though, the resistor is not dissipating any power as heat, and the voltage source is not delivering any power.
Let's now look at some point in the middle - let's say, 6A through the loop. The resistor is dissipating 6A * 6A * 1Ohm = 36W, and has a voltage drop of 6A * 1Ohm = 6V. The motor is delivering 0.01Nm/A * 6A = 0.06Nm, and has a voltage drop of 12V - 6V = 6V, so is spinning at 6V/(0.01V/(rad/s)) = 600 rad/s. As such, the motor is delivering a power of 6A * 6V = 36W. Just to double-check our numbers, 0.06Nm * 600 rad/s = 36W - we do not need to know both what the motor is doing and what its voltage and current are. The source is delivering 36W + 36W = 72W.
As a final exercise in this idealized system, let's determine the peak power the motor can deliver, assuming steady state conditions. Fortunately, we learn a couple interesting things on the way. Unfortunately, we need to do some calculus.
P_motor = V_motor * I_system
= (V_source - V_res) * I_system
= V_source * I_system - V_res * I_system
= V_source * I_system - I_system^2 * R_res
The power delivered by the motor equals the power delivered by the source, less resistive losses. This makes sense - power has to come from and go somewhere in a steady state system - but note for later that it doesn't matter where in the system the resistive losses are.
Continuing, the peak power is where either slightly more current or slightly less delivers lower power - the slope of the power-current curve is positive towards lower current, and negative towards more current, so we want the slope to equal zero, or in other words we need a point such that the derivative of power with respect to current equals zero.
dP_motor / dI_system = V_source - 2 * I_system * R_res = 0
V_source = 2 * I_system_motpeak * R_res
I_system_motpeak = V_source / R_res / 2 = 12V / 1Ohm / 2 = 6A
P_motor_peak = 12V * 6A - 6A * 6A * 1Ohm = 72W - 36W = 36W
It's like I planned this or something. Another important takeaway: the voltage source does not care whether the motor is a motor delivering some amount of torque at some speed, or a black box with a resistor appropriately sized to have some voltage drop at some current.
Now let's look at this in conjunction with a battery, which conveniently can be modeled as a voltage source in series with a resistor.
The Tesla Model S P90D has a 96S74P battery pack rated to 90kWh capacity. The cells can be charged to 4.3V, the internal resistance is on the order of 0.035Ohm, the wire bonds connecting the cells to the busbars in the modules can tolerate about 35A each without breaking, and where applicable let's assume the motor and inverter are each 95% efficient. The wire bonds themselves (1 on each end of each cell) are about 0.007Ohm. All busbar, pack fuse, connector, etc. resistances are assumed to be zero (and are in fact pretty close to negligible).
V_max = 4.3V * 96 = 412.8V
I_max = 35A * 74 = 2590A (note: this would blow their pack-level fuses in well under a second)
R_cells = 0.035Ohm / 74 * 96 = 0.045Ohm
R_pack = 0.049Ohm / 74 * 96 = 0.064Ohm
The peak power the chemical reactions in the cells are allowed to deliver is 412.8V * 2590A = 1070kW = 1434HP. The resistive losses within the cells are 2590A ^2 * 0.045Ohm = 302kW (405HP), and the resistive losses within the pack as a whole are 2590A ^2 * 0.064Ohm = 430kW.
As such, the pack can deliver a peak power of 640kW = 860HP.
It is extremely important to note that there is still a lot of current headroom based on the voltage and resistance of the pack - if you short across the pack terminals with closed contactors, you will momentarily see about
412.8V / 0.064Ohm = 6450A
before the wire bonds and pack-level fuse all blow and no more current flows. This means that we can deliver maximum current until the voltage drop across the inverter equals
412.8 V - 2590A * 0.064Ohm = 247 V
If we ask for the maximum current given an arbitrary voltage across the inverter terminals (assuming the only resistance in the system is in the pack):
I_inverter = I_system = (V_max - V_inverter) / R_pack
P_inverter = I_inverter * V_inverter
= V_max * V_inverter / R_pack - V_inverter ^2 / R_pack
If we then solve for the maximum power the inverter can take from the system (and for the sake of math I can actually demonstrate in a forum post, let's assume that corresponds to both peak inverter power delivery and peak motor power delivery), we take the derivative of inverter power with respect to inverter voltage:
dP_inverter/dV_inverter = V_max / R_pack - 2 * V_inverter / R_pack = 0
V_max / R_pack = 2 * V_inverter / R_pack
V_inverter = V_max / 2 = 412.8 V /2 = 206.4 V
Note that this is a lower voltage than the boundary condition imposed by allowable motor current! We therefore have two options for our peak power: 206.4V * 2590A or 247V * 2590A. The latter is significantly higher value, at 640kW or 858HP.
Side note: in this particular case, we can maintain our peak system torque more than halfway through the motor's speed profile. It's not quite that impressive in the real world, but there is a range of speeds at which a real motor can deliver peak allowable torque, and at max allowable speed there is a range of deliverable torques. These ranges are defined by thermal and mechanical limits respectively, and the motor constant really doesn't come into play.
Now, inverter and motor efficiencies both being 95% means the motor actually delivers a peak power of 858HP * 0.95 * 0.95 = 774HP. Tesla claims 762HP for the P90D, which could easily be accounted for by some resistance in busbars, connectors, etc., my assumed values being slightly off, and so forth.
However, this does not account for drivetrain and tire rolling efficiency - which we can reasonably take as 90% and 97% efficient respectively, for a power at the road of 676HP.
The number four thing you need to keep in mind about the electric powertrain is that battery properties are not constant.
Open-circuit (no-current) voltage drops as the state of charge decreases. Real-world capacity decreases as the rate of current draw increases. Capacity drops and internal resistance rises as cells are subjected to charge/discharge cycles, and these changes occur significantly faster during higher-current and low-temperature charge and discharge cycles.
What this means for the average consumer is that supercharging is not something you want to do all the time, you should treat your electric vehicle with care, and understand that your battery will likely be a significant influence on long-term resale value and maintenance costs.
The number five thing to keep in mind about the electric powertrain is that speed is problematic.
Back in section 2, I mentioned that air resistance is proportional to the square of speed, and at highway speeds is the main force acting against a vehicle's motion. Remember that power is force times speed - which means that, for all intents and purposes, the power a vehicle requires to maintain speed is proportional to the cube of speed - this means that your engine needs to put out 8 times as much power at 150mph as at 75mph.
Let's take an electric vehicle that has a 90kWh battery pack and takes 10kW to go 60mph. It follows that said vehicle can go 540 miles on a charge! As long as it doesn't speed up, slow down, go up a hill, run the air conditioner, run the heater, etc.
If instead the driver wanted to go the speed of traffic of 75mph, the required power is 19.5kW and the range drops to (75mph * 90kWh / 19.5kW) = 346mi.
Now let's say we're on the Autobahn. We're going 155mph in this car, requiring 172.4kW for a range of... 81 miles. If this is the mechanical top speed in a gear, a peak motor power of around 560kW is delivered around 40mph.
One more case: we pick a gear ratio that gives us peak power of 760HP=567kW at top speed. Our top speed is then 230.5mph, and we can go 37 miles. (Note that, because peak power drops as state of charge drops, we won't be able to reach that speed, nor can we maintain any top speed).
That particular case isn't very realistic, though. With a one-gear system, and assuming peak power is at about 95% of peak current, peak current would be maintained from 0-200mph - which has severe thermal implications - and the motor would be delivering about 50HP to accelerate at 40mph, thus making a Geo Metro seem sprightly.
So, why not use a multi-gear system? Efficiency. Specifically, a shifting mechanism costs about 5% drivetrain efficiency, which loosely translates to 15 miles in a 300-mile-range car - and range anxiety is one of the primary factors slowing the growth of the electric car market. The average passenger car stays below 80mph for upwards of 95% of its mileage, and because electric cars are incredibly fast in this range, manufacturers emphasize that instead (while ensuring that the Autobahn can be driven at appropriate speeds).
The following users liked this post:
daveo4porsche (08-10-2019)
04-12-2017, 11:58 AM
#2
so, thinking outside the box... and getting the general gist of your post (but without necessarily completely understanding all of the ramifications), what motor & battery pack could I use in my IMS dead 2002 996? I'd forgo the inefficiencies and want to hook it straight up to the existing trannie. Reason for that is so I'd keep the car intact should it all go to **** and be able to put a motor back in there...
Next question would be whether PCA would allow me to track the car...
Next question would be whether PCA would allow me to track the car...
04-12-2017, 06:54 PM
#3
Instructor
Thread Starter
That's a whole can of worms. There's a thread here that briefly discusses the idea, but effectively it costs as much as a new engine with worse performance and almost no range.
For the 911 specifically, there isn't a battery pack with appreciable energy (more than about 20kWh) that will fit. You're going to be designing your own system. I'd suggest using the Tesla Model S modules at about 5kWh each; you might be able to fit 6 of them, figure $10000 or so at best.
As far as the rest of the system goes, your best bet is going to be something like this for about $7000 if you don't spend years on it. It's not going to get much cheaper, and you're still limited around 80hp. Alternatively you can find a motor and inverter from a Prius, and somehow make everything fit and talk to what it needs to.
6. If a space-constrained vehicle isn't packaged around an electric powertrain, it's going to be a ton of work to retrofit one in.
For the 911 specifically, there isn't a battery pack with appreciable energy (more than about 20kWh) that will fit. You're going to be designing your own system. I'd suggest using the Tesla Model S modules at about 5kWh each; you might be able to fit 6 of them, figure $10000 or so at best.
As far as the rest of the system goes, your best bet is going to be something like this for about $7000 if you don't spend years on it. It's not going to get much cheaper, and you're still limited around 80hp. Alternatively you can find a motor and inverter from a Prius, and somehow make everything fit and talk to what it needs to.
6. If a space-constrained vehicle isn't packaged around an electric powertrain, it's going to be a ton of work to retrofit one in.
04-12-2017, 07:13 PM
#4
Gnochi - I was a mathematics minor in college but unfortunately, not an electrical engineer. Still, I understand most what you are saying. But to what end?
1. Do you work for Tesla and are shaking in your boots with the Mission E?
2. Do you work for a company working on the Mission E and are wondering how will we ever catch up with Tesla, or the electric powered cars soon to be released by BMW, Audi, Ferrari ??, and Volvo? Not to mention Ford and GM? And all those cars will probably cost considerably less than the Mission E and the Teslas.
3. OK, so much for my trying to find out who you work for
4. I am not in the market for an electric car so all I hear about electric powered cars is they are wickedly fast and silent, and have considerably fewer moving parts and components than gas powered motors.
5. The issues often mentioned as deal breakers for a lot of people are range and recharging times.
6. Again, I am not quite clear on the purpose of your post unless it was to serve as a primer of sorts on the fundamental physics of electric cars. For that we all thank you, even if we don't understand all the physics.
7. So, as an expert in the electric car industry, consider this. I just went on a 240 mile round trip to have lunch with a dear friend of mine and if I was in the market for an electric car, which one should I buy? I should mention that I typically like to drive above the speed limit (in Texas that can be up to 85mph on some highways) and where we met for lunch has no car recharging facilities at all.
8. We have plenty of sunshine and heat in Texas but I am assuming that solar is a pretty bad recharging source for high performance electric cars.
1. Do you work for Tesla and are shaking in your boots with the Mission E?
2. Do you work for a company working on the Mission E and are wondering how will we ever catch up with Tesla, or the electric powered cars soon to be released by BMW, Audi, Ferrari ??, and Volvo? Not to mention Ford and GM? And all those cars will probably cost considerably less than the Mission E and the Teslas.
3. OK, so much for my trying to find out who you work for
4. I am not in the market for an electric car so all I hear about electric powered cars is they are wickedly fast and silent, and have considerably fewer moving parts and components than gas powered motors.
5. The issues often mentioned as deal breakers for a lot of people are range and recharging times.
6. Again, I am not quite clear on the purpose of your post unless it was to serve as a primer of sorts on the fundamental physics of electric cars. For that we all thank you, even if we don't understand all the physics.
7. So, as an expert in the electric car industry, consider this. I just went on a 240 mile round trip to have lunch with a dear friend of mine and if I was in the market for an electric car, which one should I buy? I should mention that I typically like to drive above the speed limit (in Texas that can be up to 85mph on some highways) and where we met for lunch has no car recharging facilities at all.
8. We have plenty of sunshine and heat in Texas but I am assuming that solar is a pretty bad recharging source for high performance electric cars.
04-12-2017, 08:13 PM
#5
Instructor
Thread Starter
1) I'm not at Tesla (nor have I signed any NDAs with them, which is why I'm allowed to talk about their pack =D), and I'm not that worried about the Mission E. There's a lot of room in the luxury and performance EV space.
2) I've met and talked to Involved People through my job, though I'm not involved with Mission E parts, designs, etc. in any way. I'm not at liberty to discuss specifics on what I know, but it's going to be exciting.
Again, there's a lot of room in the EV market, and the more people getting involved, the better - technology will improve faster and prices will drop on the technology. I'm not that impressed by BMW's offerings - lower performance than they should be able to deliver, also somewhat biased - and it sounds like Audi will be using the unified Porsche/Audi/VW platform. Ferrari will always be their own thing. The Fords and GMs of the world absolutely have a market position! It's not going to be exceptional in any category, but it'll be pretty good for getting around - like a car in their price range should.
3) Lol =D
4) They have peak torque starting from 0 RPM, and usually hit peak power towards the end of a 0-60 cycle. They tend to be traction limited until 20mph or so, too, so yes, wicked fast is an excellent descriptor at the speeds most people tend to drive their cars most of the time. And yes, they're unnervingly silent. As a consequence, though, there are a lot of different factors to play around with to shape the sound as much as you might desire. The best article I've ever seen on the topic is one posted a few months ago by Lucid Motors.
Anecdotally, the designers of the Model S gearbox and noise/vibration/harshness teams worked together to ensure that you could hear gear whine in the cabin purely for aesthetics.
Fewer moving parts means that the motor can be significantly smaller, given adequate cooling, and can be manufactured at lower cost (the material costs will be comparable). ICE (internal combustion engine) powertrains are fairly well understood, and tooling, expertise, and supply chains are robust and plentiful; I expect it'll be a few decades before mechanic service for EV powertrains is comparable to the current state of ICE service. There is a lot less wear and tear on the complicated pieces of EV powertrains, though, and almost everything can be replaced as a unit.
5) While things are getting better in that respect, it's pretty slow improvement. Top-of-the-line batteries gain about 2-3% energy density per year; the only way to get more range quickly is to package more cells. This has other packaging consequences, good and bad, so manufacturers try to strike a balance based on their market.
Charging gets more complicated. For example, the largest supercharging connector that's reasonable to handle can carry about 300A at up to 1000V, for an absolute maximum charging power of 300kW, though lower than that depending on the vehicle's voltage. Meanwhile, from a standard 120V plug you can theoretically pull a maximum of 120V * 15A = 1.8kW, but should limit to 1.5kW. If you install a home charging station, you should be able to get up to about 10kW.
Just because the charging station can deliver power doesn't mean that the battery can charge that quickly without damage. Right now Tesla's supercharging stations allow 120kW per car, but only up to about 80% state of charge, and that does cause measurable capacity degradation after a few 10s of cycles. Slow charging (on the order of 10kW) is good for thousands of cycles.
6) That's what I was going for
It's not a trivial topic, and there are a lot of areas that just ask for confusion when compared with ICE standards, so I wanted to give some rules of thumb that will help people decode advertising, performance metrics, etc.
7) Tesla Model S, at least right now, depending on the general supercharging network in your area. Unfortunately the infrastructure isn't sufficient and the technology isn't quite there that an EV can be the sole vehicle for every household. Road trips are problematic in general; there will likely be a market for ICE rentals long into the future.
8) Yes and no. Solar without storage isn't useful; what happens when you need to charge your car and it snowed / there was a hurricane / etc.? With sufficient storage capacity, and a large enough solar panel array, that could be a decent option for a home setup or an off-grid supercharging station. When the sun is shining, and with a large enough field, you could presumably fill the storage capacity while charging a full fleet of cars.
2) I've met and talked to Involved People through my job, though I'm not involved with Mission E parts, designs, etc. in any way. I'm not at liberty to discuss specifics on what I know, but it's going to be exciting.
Again, there's a lot of room in the EV market, and the more people getting involved, the better - technology will improve faster and prices will drop on the technology. I'm not that impressed by BMW's offerings - lower performance than they should be able to deliver, also somewhat biased - and it sounds like Audi will be using the unified Porsche/Audi/VW platform. Ferrari will always be their own thing. The Fords and GMs of the world absolutely have a market position! It's not going to be exceptional in any category, but it'll be pretty good for getting around - like a car in their price range should.
3) Lol =D
4) They have peak torque starting from 0 RPM, and usually hit peak power towards the end of a 0-60 cycle. They tend to be traction limited until 20mph or so, too, so yes, wicked fast is an excellent descriptor at the speeds most people tend to drive their cars most of the time. And yes, they're unnervingly silent. As a consequence, though, there are a lot of different factors to play around with to shape the sound as much as you might desire. The best article I've ever seen on the topic is one posted a few months ago by Lucid Motors.
Anecdotally, the designers of the Model S gearbox and noise/vibration/harshness teams worked together to ensure that you could hear gear whine in the cabin purely for aesthetics.
Fewer moving parts means that the motor can be significantly smaller, given adequate cooling, and can be manufactured at lower cost (the material costs will be comparable). ICE (internal combustion engine) powertrains are fairly well understood, and tooling, expertise, and supply chains are robust and plentiful; I expect it'll be a few decades before mechanic service for EV powertrains is comparable to the current state of ICE service. There is a lot less wear and tear on the complicated pieces of EV powertrains, though, and almost everything can be replaced as a unit.
5) While things are getting better in that respect, it's pretty slow improvement. Top-of-the-line batteries gain about 2-3% energy density per year; the only way to get more range quickly is to package more cells. This has other packaging consequences, good and bad, so manufacturers try to strike a balance based on their market.
Charging gets more complicated. For example, the largest supercharging connector that's reasonable to handle can carry about 300A at up to 1000V, for an absolute maximum charging power of 300kW, though lower than that depending on the vehicle's voltage. Meanwhile, from a standard 120V plug you can theoretically pull a maximum of 120V * 15A = 1.8kW, but should limit to 1.5kW. If you install a home charging station, you should be able to get up to about 10kW.
Just because the charging station can deliver power doesn't mean that the battery can charge that quickly without damage. Right now Tesla's supercharging stations allow 120kW per car, but only up to about 80% state of charge, and that does cause measurable capacity degradation after a few 10s of cycles. Slow charging (on the order of 10kW) is good for thousands of cycles.
6) That's what I was going for
It's not a trivial topic, and there are a lot of areas that just ask for confusion when compared with ICE standards, so I wanted to give some rules of thumb that will help people decode advertising, performance metrics, etc.7) Tesla Model S, at least right now, depending on the general supercharging network in your area. Unfortunately the infrastructure isn't sufficient and the technology isn't quite there that an EV can be the sole vehicle for every household. Road trips are problematic in general; there will likely be a market for ICE rentals long into the future.
8) Yes and no. Solar without storage isn't useful; what happens when you need to charge your car and it snowed / there was a hurricane / etc.? With sufficient storage capacity, and a large enough solar panel array, that could be a decent option for a home setup or an off-grid supercharging station. When the sun is shining, and with a large enough field, you could presumably fill the storage capacity while charging a full fleet of cars.
04-12-2017, 08:34 PM
#6
Instructor
Thread Starter
Oh, one other interesting difference between ICE and EV powertrains: accessories.
Every modern passenger car has both heating and a vacuum assist on the brake system. In an ICE car, the interior heating is sourced from the heat of combustion. There is a heat exchange unit between the engine coolant and the cabin airflow.
Without a vacuum booster, you probably wouldn't be able to stop your car effectively in anything resembling a decent stopping distance. The vacuum booster takes advantage of hydraulics to push against the brake calipers much harder than you are physically capable of, and is enabled by the manifold pressure in a normally aspirated ICE.
In an EV, there's no guarantee that the coolant will be warm enough to noticeably heat the cabin. Instead, you have to spend some of the battery system's power output on an electric heater in the cabin (the cabin does warm up much faster in winter in an EV than in an ICE). You also need a vacuum pump specifically for the brake booster, and for efficiency related reasons it needs to be its own device, instead of being tied to the crankshaft with a serpentine belt (as forced-induction engines do).
This also applies to the AC compressor, coolant pumps, power steering pumps, etc. You do get to skip the alternator and use a DC-DC converter to get from the high voltage (nominal 400-ish volt) pack to your 12V communications and such instead.
Every modern passenger car has both heating and a vacuum assist on the brake system. In an ICE car, the interior heating is sourced from the heat of combustion. There is a heat exchange unit between the engine coolant and the cabin airflow.
Without a vacuum booster, you probably wouldn't be able to stop your car effectively in anything resembling a decent stopping distance. The vacuum booster takes advantage of hydraulics to push against the brake calipers much harder than you are physically capable of, and is enabled by the manifold pressure in a normally aspirated ICE.
In an EV, there's no guarantee that the coolant will be warm enough to noticeably heat the cabin. Instead, you have to spend some of the battery system's power output on an electric heater in the cabin (the cabin does warm up much faster in winter in an EV than in an ICE). You also need a vacuum pump specifically for the brake booster, and for efficiency related reasons it needs to be its own device, instead of being tied to the crankshaft with a serpentine belt (as forced-induction engines do).
This also applies to the AC compressor, coolant pumps, power steering pumps, etc. You do get to skip the alternator and use a DC-DC converter to get from the high voltage (nominal 400-ish volt) pack to your 12V communications and such instead.
04-12-2017, 11:42 PM
#7
EV motors - interesting about size and noise designs. I stumbled across an Eclass race on the Velocity channel and was amazed how quiet they were. I was also amazed that some cars ran out of power before the end of the race. I guess you can't do a quick charge when you have a pit stop.
Speed and recharging - so the degradation of car batteries is a problem the faster you charge. And unless you are finished driving for the day, no one wants to spend hours recharging the batteries.
Speed and battery drain - What about higher speeds and increased battery drain? Am I correct to assume the faster you drive the more the drain?
Tesla - I think Tesla is to be commended for producing EVs which have set the standard.... So far. But I have read with interest that Tesla has more market capitalization on the stock market than GM or Ford. Which makes no economic sense since GM sold 10 million cars, Ford sold 6.6 million cars and Tesla sold an anemic 76 thousand cars. GM has a market capitalization of about $5,500 per car, Ford about $6,600 per car, and Tesla about $673,000 per car!!!! I don't know if I would consider a Tesla because despite being the hottest stock on the market, they have lost millions and have captured .2% of the market. In my opinion their business model will not last long and in the past two weeks several investment banks have urged investors to "sell." I am aware stocks are bought on future prospects, not the past, but when GM, Ford and the Japanese makers introduce solid E-cars at considerably lower prices than Tesla, how will Tesla survive ?
Service - You mentioned EVs will allow for unit replacement should components problems arise. That's interesting. I imagine repair costs will still be expensive. Any thoughts on reliability? Can an EV last 100,000 miles with the original batteries?
BRAKES - Are the KER systems on F1 cars too expensive for the public and it is feasible to think a racing KER system could recharge batteries through braking inertia?
Heating, AC and other systems - Wow. What you posted makes sense but I had no idea those challenges must be approached differently with EVs.
Good stuff, thanks.
Speed and recharging - so the degradation of car batteries is a problem the faster you charge. And unless you are finished driving for the day, no one wants to spend hours recharging the batteries.
Speed and battery drain - What about higher speeds and increased battery drain? Am I correct to assume the faster you drive the more the drain?
Tesla - I think Tesla is to be commended for producing EVs which have set the standard.... So far. But I have read with interest that Tesla has more market capitalization on the stock market than GM or Ford. Which makes no economic sense since GM sold 10 million cars, Ford sold 6.6 million cars and Tesla sold an anemic 76 thousand cars. GM has a market capitalization of about $5,500 per car, Ford about $6,600 per car, and Tesla about $673,000 per car!!!! I don't know if I would consider a Tesla because despite being the hottest stock on the market, they have lost millions and have captured .2% of the market. In my opinion their business model will not last long and in the past two weeks several investment banks have urged investors to "sell." I am aware stocks are bought on future prospects, not the past, but when GM, Ford and the Japanese makers introduce solid E-cars at considerably lower prices than Tesla, how will Tesla survive ?
Service - You mentioned EVs will allow for unit replacement should components problems arise. That's interesting. I imagine repair costs will still be expensive. Any thoughts on reliability? Can an EV last 100,000 miles with the original batteries?
BRAKES - Are the KER systems on F1 cars too expensive for the public and it is feasible to think a racing KER system could recharge batteries through braking inertia?
Heating, AC and other systems - Wow. What you posted makes sense but I had no idea those challenges must be approached differently with EVs.
Good stuff, thanks.
Trending Topics
04-13-2017, 12:04 AM
#8
Instructor
Thread Starter
1) Yeah, until the 2018 season you'll swap batteries halfway through the race. They're changing things up a bit after that.
2) Correct. Commuting is probably fine, long road trips probably aren't. One of the biggest things being developed now are supercharging-tolerant cell chemistries, where you might see 5% capacity degradation after 10k cycles. Last I heard that should be available 2020-ish if things improve at the current rate.
3) Yes, higher speeds require more power from the battery, which increases degradation rate. Keeping the cells cooler (below a maximum of about 45degC) helps somewhat.
4) I'm with you on that one, in terms of market capitalization. Stock markets are weird. I'll stick with seinsible electromechanical widgets instead. As far as I can tell Tesla is pushing the technology as much as they can when the markets are in their favor, which will let them weather the times the markets aren't.
5) To add some clarity, I'm talking about things like motors, inverters, battery modules, etc. There are a lot of sensitive components packaged together in a small area with coolant, wire bonds, busbars, epoxy, etc. all holding it together. At the manufacturer level, some refurbishment might be doable - I've replaced bits and pieces on assorted prototype packs - but it's not the sort of thing a manufacturer would want to let someone do without a release of liability and a acknowledgement of fully-voided systemwide warranty.
As far as 100,000 miles with original batteries go, if you baby the vehicle you might see 5% capacity degradation, and if you drive it like you stole it, you might see 40% degradation. I foresee driving and charging records being more valuable than maintenance records over the next couple decades.
That said, a 60% capacity pack is still useful for stationary storage purposes, and a pack sale could cover a decent fraction of the cost of a replacement pack with several years better technology.
6) Thanks to the magic of the electric powertrain, you don't even need a special system to do this, just a few additional lines of code. If you take a dc motor, connect the input and output to a lightbulb, and spin the axle, the lightbulb turns on. It's slightly more involved with an AC motor and inverter, but the same principle applies.
A colleague of mine charged his batteries at the top of Pike's Peak and almost burned out his brakes on the way down, because the manufacturer had sized the brakes to assume regenerative braking on long downhills, and didn't think of the case where the battery is already fully charged. There are a lot of ways to waste this extra power generation and they hadn't enabled any of them (running the AC and heater at the same time, for example). They fixed that in a software patch a couple weeks later.
Note that this doesn't work on, for example, a Boosted longboard. You *only* have regenerative braking, and if the battery is fully charged, you're going to need a different way to slow down (and the resistors in the balancing circuitry aren't large enough to bleed off sufficient power). Jackets make decent parachutes without losing control, by the way.
7) That was one of my big "wait, what?!" moments a couple years ago, because it's really not something people would ever think about!
I find this stuff interesting and fun and I can talk about it for hours with a willing audience
2) Correct. Commuting is probably fine, long road trips probably aren't. One of the biggest things being developed now are supercharging-tolerant cell chemistries, where you might see 5% capacity degradation after 10k cycles. Last I heard that should be available 2020-ish if things improve at the current rate.
3) Yes, higher speeds require more power from the battery, which increases degradation rate. Keeping the cells cooler (below a maximum of about 45degC) helps somewhat.
4) I'm with you on that one, in terms of market capitalization. Stock markets are weird. I'll stick with seinsible electromechanical widgets instead. As far as I can tell Tesla is pushing the technology as much as they can when the markets are in their favor, which will let them weather the times the markets aren't.
5) To add some clarity, I'm talking about things like motors, inverters, battery modules, etc. There are a lot of sensitive components packaged together in a small area with coolant, wire bonds, busbars, epoxy, etc. all holding it together. At the manufacturer level, some refurbishment might be doable - I've replaced bits and pieces on assorted prototype packs - but it's not the sort of thing a manufacturer would want to let someone do without a release of liability and a acknowledgement of fully-voided systemwide warranty.
As far as 100,000 miles with original batteries go, if you baby the vehicle you might see 5% capacity degradation, and if you drive it like you stole it, you might see 40% degradation. I foresee driving and charging records being more valuable than maintenance records over the next couple decades.
That said, a 60% capacity pack is still useful for stationary storage purposes, and a pack sale could cover a decent fraction of the cost of a replacement pack with several years better technology.
6) Thanks to the magic of the electric powertrain, you don't even need a special system to do this, just a few additional lines of code. If you take a dc motor, connect the input and output to a lightbulb, and spin the axle, the lightbulb turns on. It's slightly more involved with an AC motor and inverter, but the same principle applies.
A colleague of mine charged his batteries at the top of Pike's Peak and almost burned out his brakes on the way down, because the manufacturer had sized the brakes to assume regenerative braking on long downhills, and didn't think of the case where the battery is already fully charged. There are a lot of ways to waste this extra power generation and they hadn't enabled any of them (running the AC and heater at the same time, for example). They fixed that in a software patch a couple weeks later.
Note that this doesn't work on, for example, a Boosted longboard. You *only* have regenerative braking, and if the battery is fully charged, you're going to need a different way to slow down (and the resistors in the balancing circuitry aren't large enough to bleed off sufficient power). Jackets make decent parachutes without losing control, by the way.
7) That was one of my big "wait, what?!" moments a couple years ago, because it's really not something people would ever think about!
I find this stuff interesting and fun and I can talk about it for hours with a willing audience
04-13-2017, 10:24 AM
#9
GNOCHI wrote:
1) Yeah, until the 2018 season you'll swap batteries halfway through the race. They're changing things up a bit after that.
You know I never thought I was a "sound" guy when it came to Porsche's or racing but those e-class cars need some sort of engine noise. Sound adds far more enjoyment to high speed than I thought.
4) I'm with you on that one, in terms of market capitalization. Stock markets are weird. I'll stick with seinsible electromechanical widgets instead. As far as I can tell Tesla is pushing the technology as much as they can when the markets are in their favor, which will let them weather the times the markets aren't.
This may be hard to speculate about, but say someone buys a Tesla and then Tesla fails, or better yet, there is not a Telsa dealership in your town just like there is not a Porsche dealership where I live. Under warranty you can drive to the nearest dealership. After the warranty expires, what are your thoughts on how will EVs be maintained? Do you expect "indies" will develop that can do all the things to keep a car running or will dealerships be the only option, even after warranty.
5) ... but it's not the sort of thing a manufacturer would want to let someone do without a release of liability and a acknowledgement of fully-voided systemwide warranty.
Are you saying an untrained, fumbled-fingered mechanic can cause some major electrical damage with an errant screwdriver? What else is new
As far as 100,000 miles with original batteries go, if you baby the vehicle you might see 5% capacity degradation, and if you drive it like you stole it, you might see 40% degradation. I foresee driving and charging records being more valuable than maintenance records over the next couple decades.
I guess there will be 1,000,000 mile Mission E cars working on their 15th set of batteries? This might upset the collectors out there
That said, a 60% capacity pack is still useful for stationary storage purposes, and a pack sale could cover a decent fraction of the cost of a replacement pack with several years better technology.
Pack sale?
And just to be sure I understand you, with only 60% capacity after five years of driving it like I stole it, the range of an EV would be reduced 40% or so.
6) ... A colleague of mine charged his batteries at the top of Pike's Peak and almost burned out his brakes on the way down, because the manufacturer had sized the brakes to assume regenerative braking on long downhills, and didn't think of the case where the battery is already fully charged. There are a lot of ways to waste this extra power generation and they hadn't enabled any of them (running the AC and heater at the same time, for example). They fixed that in a software patch a couple weeks later.
Fascinating and I can't blame an EV developer for not thinking of this scenario
Note that this doesn't work on, for example, a Boosted longboard. You *only* have regenerative braking, and if the battery is fully charged, you're going to need a different way to slow down (and the resistors in the balancing circuitry aren't large enough to bleed off sufficient power). Jackets make decent parachutes without losing control, by the way.
Again, wow. No wonder F1 drivers are always fiddling with braking adjustments and you know those KER systems have to be over-engineered for the demands of F1 tracks
7) That was one of my big "wait, what?!" moments a couple years ago, because it's really not something people would ever think about!
Which certainly begs the comment that ICE cars utilize every available source of energy and assistance that the mechanical systems develop. I am sure EVs will get there too, but in ways that we don't normally think about.
Thanks again...
1) Yeah, until the 2018 season you'll swap batteries halfway through the race. They're changing things up a bit after that.
You know I never thought I was a "sound" guy when it came to Porsche's or racing but those e-class cars need some sort of engine noise. Sound adds far more enjoyment to high speed than I thought.
4) I'm with you on that one, in terms of market capitalization. Stock markets are weird. I'll stick with seinsible electromechanical widgets instead. As far as I can tell Tesla is pushing the technology as much as they can when the markets are in their favor, which will let them weather the times the markets aren't.
This may be hard to speculate about, but say someone buys a Tesla and then Tesla fails, or better yet, there is not a Telsa dealership in your town just like there is not a Porsche dealership where I live. Under warranty you can drive to the nearest dealership. After the warranty expires, what are your thoughts on how will EVs be maintained? Do you expect "indies" will develop that can do all the things to keep a car running or will dealerships be the only option, even after warranty.
5) ... but it's not the sort of thing a manufacturer would want to let someone do without a release of liability and a acknowledgement of fully-voided systemwide warranty.
Are you saying an untrained, fumbled-fingered mechanic can cause some major electrical damage with an errant screwdriver? What else is new
As far as 100,000 miles with original batteries go, if you baby the vehicle you might see 5% capacity degradation, and if you drive it like you stole it, you might see 40% degradation. I foresee driving and charging records being more valuable than maintenance records over the next couple decades.
I guess there will be 1,000,000 mile Mission E cars working on their 15th set of batteries? This might upset the collectors out there
That said, a 60% capacity pack is still useful for stationary storage purposes, and a pack sale could cover a decent fraction of the cost of a replacement pack with several years better technology.
Pack sale?
And just to be sure I understand you, with only 60% capacity after five years of driving it like I stole it, the range of an EV would be reduced 40% or so.
6) ... A colleague of mine charged his batteries at the top of Pike's Peak and almost burned out his brakes on the way down, because the manufacturer had sized the brakes to assume regenerative braking on long downhills, and didn't think of the case where the battery is already fully charged. There are a lot of ways to waste this extra power generation and they hadn't enabled any of them (running the AC and heater at the same time, for example). They fixed that in a software patch a couple weeks later.
Fascinating and I can't blame an EV developer for not thinking of this scenario
Note that this doesn't work on, for example, a Boosted longboard. You *only* have regenerative braking, and if the battery is fully charged, you're going to need a different way to slow down (and the resistors in the balancing circuitry aren't large enough to bleed off sufficient power). Jackets make decent parachutes without losing control, by the way.
Again, wow. No wonder F1 drivers are always fiddling with braking adjustments and you know those KER systems have to be over-engineered for the demands of F1 tracks
7) That was one of my big "wait, what?!" moments a couple years ago, because it's really not something people would ever think about!
Which certainly begs the comment that ICE cars utilize every available source of energy and assistance that the mechanical systems develop. I am sure EVs will get there too, but in ways that we don't normally think about.
Thanks again...
04-13-2017, 12:41 PM
#10
Rennlist Member
Thanks for a nice presentation of a lot of the less intuitive results of this new tech.
I'm a Mechanical Engineer working for big aerospace and was seriously considering stopping to create an EV startup a few years ago and again last year. Lots of very smart folks are willing to work like mad to be a part of the wave. I had 90% of a business plan with plans to build a running prototype/proof of concept, do a few ads that I thought had a high chance of going viral, and going public ASAP after that to realize a large financial gain. I'd plan to hand over the reins soon after as someone else would be better able to scale up from a startup into a real company. I stopped both times as I ended up deciding I wasn't willing to put in the work for the years it would take to find out if my venture was successful or would fizzle out. Very interesting tech to work on and staggering amounts of money are being thrown at it.
Hope you're enjoying your career!
I'm a Mechanical Engineer working for big aerospace and was seriously considering stopping to create an EV startup a few years ago and again last year. Lots of very smart folks are willing to work like mad to be a part of the wave. I had 90% of a business plan with plans to build a running prototype/proof of concept, do a few ads that I thought had a high chance of going viral, and going public ASAP after that to realize a large financial gain. I'd plan to hand over the reins soon after as someone else would be better able to scale up from a startup into a real company. I stopped both times as I ended up deciding I wasn't willing to put in the work for the years it would take to find out if my venture was successful or would fizzle out. Very interesting tech to work on and staggering amounts of money are being thrown at it.
Hope you're enjoying your career!
04-14-2017, 01:13 AM
#11
Instructor
Thread Starter
@ocgarza:
1) There are many reasons I drive a Porsche instead of an EV. This is one of them.
4) That's an interesting topic. 95% of the car could probably be done at an indy - wheel bearings, transmission rebuilds, cooling line replacement, various interior bits and bobs - but I just can't see a way for the core powertrain to be serviced outside of a dealership. Perhaps they'll encourage dealer-level service with a core credit on those elements, even after the warranty has expired?
5) That's essentially correct re: capacity loss, though that's one of the areas that EVs are weird. Re: pack sale: you might be able to sell a 5-year-old 60%-capacity pack for 25% of the cost of a new pack with new technology. Even if the energy density is no longer there for a vehicle the modules could have second life in onsite power storage applications.
7) I'd argue that on a ICE powertrain, the power drain from all of those auxiliary units is lost in the noise anyway, so it doesn't really matter how efficient they are, and not requiring an extra motor for each of them is an excellent cost (and packaging space) savings measure.
With an EV, you really do care about that 4kW AC compressor, since it cuts your 60mph range by about 25%! As such, you'll only run those elements when you need to, with only as much power as they require, instead of accepting the additional parasitic losses.
@ace37
Thank you! I love what I'm doing and I can't see myself switching fields anytime soon.
A few of our best engineers come from aerospace backgrounds; we have slightly larger tolerances and a couple orders of magnitude less paperwork, in case that sounds like something you may be interested in =D
It really is a ton of work for years - and hundreds of millions of dollars - to figure out whether any automotive startup will be worth anything. I do like staggering amounts of available funding, but I know I don't have the personality to be a company- or market-driving influence. I like making widgets work, getting things to package together nicely, and playing with fun toys.
1) There are many reasons I drive a Porsche instead of an EV. This is one of them.
4) That's an interesting topic. 95% of the car could probably be done at an indy - wheel bearings, transmission rebuilds, cooling line replacement, various interior bits and bobs - but I just can't see a way for the core powertrain to be serviced outside of a dealership. Perhaps they'll encourage dealer-level service with a core credit on those elements, even after the warranty has expired?
5) That's essentially correct re: capacity loss, though that's one of the areas that EVs are weird. Re: pack sale: you might be able to sell a 5-year-old 60%-capacity pack for 25% of the cost of a new pack with new technology. Even if the energy density is no longer there for a vehicle the modules could have second life in onsite power storage applications.
7) I'd argue that on a ICE powertrain, the power drain from all of those auxiliary units is lost in the noise anyway, so it doesn't really matter how efficient they are, and not requiring an extra motor for each of them is an excellent cost (and packaging space) savings measure.
With an EV, you really do care about that 4kW AC compressor, since it cuts your 60mph range by about 25%! As such, you'll only run those elements when you need to, with only as much power as they require, instead of accepting the additional parasitic losses.
@ace37
Thank you! I love what I'm doing and I can't see myself switching fields anytime soon.
A few of our best engineers come from aerospace backgrounds; we have slightly larger tolerances and a couple orders of magnitude less paperwork, in case that sounds like something you may be interested in =D
It really is a ton of work for years - and hundreds of millions of dollars - to figure out whether any automotive startup will be worth anything. I do like staggering amounts of available funding, but I know I don't have the personality to be a company- or market-driving influence. I like making widgets work, getting things to package together nicely, and playing with fun toys.
04-14-2017, 11:47 AM
#12
Rennlist Member
Methinks that todays liquid electrolyte lithium ion batteries and metal magnets are going to be archaic in a few years as ceramic magnet motors and plastic electrolyte lithium anode batteries will blow this stuff away.
When looking at the limitations of the liquid electrolyte batteries in terms of slow charge times, capacity loss and inherent fire risk, it makes sense for Porsche to sidestep/skip this old battery technology.
The Mission-E needs to lead, rather than follow...
When looking at the limitations of the liquid electrolyte batteries in terms of slow charge times, capacity loss and inherent fire risk, it makes sense for Porsche to sidestep/skip this old battery technology.
The Mission-E needs to lead, rather than follow...
04-14-2017, 05:29 PM
#13
Rennlist Member
@ace37
Thank you! I love what I'm doing and I can't see myself switching fields anytime soon.
A few of our best engineers come from aerospace backgrounds; we have slightly larger tolerances and a couple orders of magnitude less paperwork, in case that sounds like something you may be interested in =D
It really is a ton of work for years - and hundreds of millions of dollars - to figure out whether any automotive startup will be worth anything. I do like staggering amounts of available funding, but I know I don't have the personality to be a company- or market-driving influence. I like making widgets work, getting things to package together nicely, and playing with fun toys.
Thank you! I love what I'm doing and I can't see myself switching fields anytime soon.
A few of our best engineers come from aerospace backgrounds; we have slightly larger tolerances and a couple orders of magnitude less paperwork, in case that sounds like something you may be interested in =D
It really is a ton of work for years - and hundreds of millions of dollars - to figure out whether any automotive startup will be worth anything. I do like staggering amounts of available funding, but I know I don't have the personality to be a company- or market-driving influence. I like making widgets work, getting things to package together nicely, and playing with fun toys.
What are the typical failure modes you take a real close look at with the induction motors, just things like bearing wear and overheating?
And is ensuring good thermal management across a broad range of conditions is one of the biggest workloads for a technical group?
04-18-2017, 02:37 AM
#14
Instructor
Thread Starter
Busy weekend led to slow response. Sorry!
@928 GT R
In a few years, quite possibly, but I don't think the battery technology at least is there quite yet, and the cost is through the roof. The solid state cells I've seen and played with have excellent energy density and efficiency, but couldn't push much current at all. There are a lot of university teams researching improvements, and the big cell manufacturers are doing what they can on their end as well.
I can't speak much on the magnet front; it looks like there's an industry-wide trend to move in the direction of induction motors due to the cost and weight of magnets at all.
These days the trick is to make your EV technology-agnostic so you can switch cell chemistries/motor technologies/etc as they come along without needing to do full tearups of the relevant systems. Porsche, Tesla, Lucid - all of the tech-forefront companies will likely be able to swap relatively quickly as developments become available. Depending on price and timing they may be able to swap partway through a production year.
@ace37
That does sound like the perfect job! That's what I'm doing now, only it's 60 to 80 hours a week (and I've worked a handful of 100s in the past few months).
Aside from insert-interface-here-leaking, every failure I've seen has boiled down to overheating in some respect - thermal cracking from hot spots, melted windings, insulation failures, and the like. I just go by what I overhear, though; I don't spend a lot of time in the dyno lab, since I can use a resistor bank instead.
Thermal management is probably the biggest workload for a given group, yeah. It's the second biggest driver of package (after performance) and is the third biggest limiter of performance (after connectors [driven by package, then thermals], contactor limits [driven by availability, then package, then thermals], and fuse limits [driven by package, thermals, and controls]).
If you're getting the sense that it's a rat's nest of dependencies, you're correct, and that's why there is no best answer for every situation, only a collectively optimal answer for some subset of situations.
@928 GT R
In a few years, quite possibly, but I don't think the battery technology at least is there quite yet, and the cost is through the roof. The solid state cells I've seen and played with have excellent energy density and efficiency, but couldn't push much current at all. There are a lot of university teams researching improvements, and the big cell manufacturers are doing what they can on their end as well.
I can't speak much on the magnet front; it looks like there's an industry-wide trend to move in the direction of induction motors due to the cost and weight of magnets at all.
These days the trick is to make your EV technology-agnostic so you can switch cell chemistries/motor technologies/etc as they come along without needing to do full tearups of the relevant systems. Porsche, Tesla, Lucid - all of the tech-forefront companies will likely be able to swap relatively quickly as developments become available. Depending on price and timing they may be able to swap partway through a production year.
@ace37
That does sound like the perfect job! That's what I'm doing now, only it's 60 to 80 hours a week (and I've worked a handful of 100s in the past few months).
Aside from insert-interface-here-leaking, every failure I've seen has boiled down to overheating in some respect - thermal cracking from hot spots, melted windings, insulation failures, and the like. I just go by what I overhear, though; I don't spend a lot of time in the dyno lab, since I can use a resistor bank instead.
Thermal management is probably the biggest workload for a given group, yeah. It's the second biggest driver of package (after performance) and is the third biggest limiter of performance (after connectors [driven by package, then thermals], contactor limits [driven by availability, then package, then thermals], and fuse limits [driven by package, thermals, and controls]).
If you're getting the sense that it's a rat's nest of dependencies, you're correct, and that's why there is no best answer for every situation, only a collectively optimal answer for some subset of situations.
