Intake to Port question?
05-22-2015, 06:17 AM
#16
Addict
Rennlist Member
Rennlist Member
Very nice, lenght seems to be in the neighborhood of famous porsche 924/944 race engines.
05-22-2015, 12:23 PM
#18
Race Car
Sorry I'm late to the party. But here's a brief bit of info.
Intake runners are all about TUNING. They are not about flow (to a degree anyway). Ideally, you'd have them taper from the choke point a little upstream from the valve all of the way to the bellmouth. But that is not very practical in very many applications at all. So most of what I've designed has kept the same diameter. The way I approach it - determine RPM for peak power, size the port according to that. Figure up to 10% taper from port diameter to valve seat, then factor in seat geometry to get the valve size. If it won't work, determine what will work (valve size wise). For example, you may need an ID on your runners of 1.70", but you can only find a 1.75" OD, 14 gauge tube, meaning you have to to deal with a 1.62" ID tube.
Once you have your ID identified, then you need to determine at what RPM peak torque would be. This will determine the overall length of the runner. For example, if you end up with a 1.62" ID tube which is from a 6800 RPM horsepower number (and adjusted for what kind of tube is available), then you want to hit peak torque somewhere around 5000 RPM, then you'll end up with a runner length (from bellmouth to valve) of about 13" (330mm).
Anyway, to get to the "smooth versus step" question. Every time you have a step, you will create a reversion wave. It would be like ending the runner right there. It will try to tune like a shorter runner intake. If you WANT it to tune there, leave a step. If you do NOT want it to tune there, you want to make the transition as gradual as you possibly can. Real world example on my 968. Ideally, I'd have a 2.5" exhaust pipe, 30" long from the collector. That would put it ahead of the torsion tube. So to run it through the back of the car, it is best to enlarge it. Previously, most of the exhaust systems that Karl built, he'd use a gradual transition to 3" right there. The last 968 he did made 236 at the wheels (a more typical number is 230). There were a few differences that would factor in, some in my favor (no PS), some in favor of the last 968 (nikasil bores). But they should be between balancing each other to giving the last 968 an edge.
This time around, instead of using a gradual opening to 3", I used an abrupt step so that it would create a reversion wave to help it make a lower pressure at the collector. There is no question that it would not flow as well as the gradual expansion. But, it would tune better. The result? 244 horsepower.
In that case, I was trying to do the opposite of what Patrick is trying to do here, in that I wanted to intentionally create a resonance at that point. But it should give a little bit of an analogy of the importance of tuning.
Intake runners are all about TUNING. They are not about flow (to a degree anyway). Ideally, you'd have them taper from the choke point a little upstream from the valve all of the way to the bellmouth. But that is not very practical in very many applications at all. So most of what I've designed has kept the same diameter. The way I approach it - determine RPM for peak power, size the port according to that. Figure up to 10% taper from port diameter to valve seat, then factor in seat geometry to get the valve size. If it won't work, determine what will work (valve size wise). For example, you may need an ID on your runners of 1.70", but you can only find a 1.75" OD, 14 gauge tube, meaning you have to to deal with a 1.62" ID tube.
Once you have your ID identified, then you need to determine at what RPM peak torque would be. This will determine the overall length of the runner. For example, if you end up with a 1.62" ID tube which is from a 6800 RPM horsepower number (and adjusted for what kind of tube is available), then you want to hit peak torque somewhere around 5000 RPM, then you'll end up with a runner length (from bellmouth to valve) of about 13" (330mm).
Anyway, to get to the "smooth versus step" question. Every time you have a step, you will create a reversion wave. It would be like ending the runner right there. It will try to tune like a shorter runner intake. If you WANT it to tune there, leave a step. If you do NOT want it to tune there, you want to make the transition as gradual as you possibly can. Real world example on my 968. Ideally, I'd have a 2.5" exhaust pipe, 30" long from the collector. That would put it ahead of the torsion tube. So to run it through the back of the car, it is best to enlarge it. Previously, most of the exhaust systems that Karl built, he'd use a gradual transition to 3" right there. The last 968 he did made 236 at the wheels (a more typical number is 230). There were a few differences that would factor in, some in my favor (no PS), some in favor of the last 968 (nikasil bores). But they should be between balancing each other to giving the last 968 an edge.
This time around, instead of using a gradual opening to 3", I used an abrupt step so that it would create a reversion wave to help it make a lower pressure at the collector. There is no question that it would not flow as well as the gradual expansion. But, it would tune better. The result? 244 horsepower.
In that case, I was trying to do the opposite of what Patrick is trying to do here, in that I wanted to intentionally create a resonance at that point. But it should give a little bit of an analogy of the importance of tuning.
05-22-2015, 02:23 PM
#20
Drifting
I remember the Rimflow valves from decades ago, I think David Vizard came up with the design. Basically there was a groove machined into the cylinder face side of the inlet valve to hinder flow back into the inlet port, and a similar ridge on the back of the exhaust valve to prevent exhaust flow back into the cylinder. There's more to cylinder filling than port/manifold V valve size.
05-22-2015, 04:23 PM
#21
05-22-2015, 06:52 PM
#22
Rennlist Member
Thread Starter
Yes, as stated, the runners are longer than many would prescribe. Hey, it may fall flat on it's face but we were going off what were advised. It was definitely a grab for more torque with such a different motor. Perhaps we're weakening a strength while strengthening a weakness. We shall see.
This from the Intake person who recommended these lengths. These are from a series of emails. So they are somewhat informal but you will get the gist. In the end we had to use 1 size down from his suggested Velocity stacks. We're 48mm I.D. which have a 94mm Bell not the 50/100m he suggested.
"Ok, some numbers, ill just type them out in no particular order as I run the calcs
Firstly your port flow seems right on the money. I calculate max required intake port flow at 240-254cfm @28". Exhaust port at 203-220cfm @ 28"
Pipemax suggests you can keep pushing the lift, and it'll keep flowing, but we knew that already
Calculated power band 5200-7300rpm
Port velocities are going to be a bit slow based on your port size at flange (would need the exact area of the port to say exactly, but based on gross dimensions its on the bigger side of just acceptable) 3rd Harmonic tuning will prob work out best for you, taking packaging into consideration (this gives you longer runners and more torque, whilst still tuning for RPM specific peak HP)runner length (from back of valve) = 317mm looking at port speed and runner length, runner ID should be 50mm. That's a narrower cross section to your port at the flange, but a bit of forming and grinding will get a smooth enough transition.
Id be getting these: http://www.jntperformance.com.au/ind...l?catId=269035
Nice taper over the length, then just 50mm ID tube/bend to the flange what youll probably find is with 2bar of boost, youll lose a bit of that (effective) diamater to boundary layer effects, in essence itll act like its a bit smaller, and your velocities will pick up a bit (which in your case is a good thing). given the radius of those bellmouth, id make the length from bellmouth lip to valve 320mm."
"Intake acoustic harmonics always go in 'tone' steps. They're all designed to 'come on song' at the same particular frequency/note (ie rpm) but the energy in the harmonic goes down as you go to shorter and shorter steps (higher octave pitch). Longer gives more energy (low note), and make therefore more torque (everywhere), with the exception that as you go really long into really low harmonics, you start having major packaging problems (500mm long runners etc), and they have horrible losses at really high flow rates. OEM make really long, curly, spiral intakes which make great torque when the engine isn't flowing much at lower rpms, but can't handle the absolute flow demands for peak flow at high rpms.
You've got two length choices, one (from memory) around 250mm, and the one I gave you at 320mm. The longer one I think in your case, has no packaging disadvantage, and might actually help get your plenum in the right spot with the tb somewhere useful. It'll produce more tq, and I can't think that with a 50mm ID runner and a nice tapered bell, that you're going to have high flow issues. As I mentioned, if anything, it'll get the intake velocity a bit quicker, and you can use that (you're nowhere near choking velocity). If it were mine, I'd pick the longer runner.
Regarding plenum... A gentle bend to the runner is fine. I'd be guessing the ports in the head are about, I dunno, 80mm long (total guess). The bellmouths are 150mm long. That gives you ~90mm of space to slot in a bit of 50mm ID gentle mandrel bend. Bellmouths should be flat in the plenum base, you want the flow in the plenum body to be as symmetric as possible. Offset runners, angled bells etc... try and avoid
Of note, those bellmouths are HUGE... 100mm bell diameter, and 50mm ID at the end. It'll look nuts! For that size bell, I hope you've got enough room for that width of plenum. Ideally the plenum should be a bit wider, but fabrication could be a bit of a ****... 4" alloy tube and donuts and bends are easy to get. 5"? Not so much! Can stretch the 4" open a few mm to get it all to fit!
Length of your old manifold is kinda irrelevant, as the new cams, the new flow rates, and the bigger (50mm) runner dia determine the length, and have no recognizable relation to the old motor/manifold. And the port size at the flange in the head, don't lose sleep over it. Can't change it, and it's not 'bad' by any measure. Just on the bigger side. 50mm runner will match to it pretty neatly, and the velocity coming down the runner won't be too disturbed by a small step or change in port profile."
"320mm from bell to valve
the 150mm bellmouth from JNT is just 150mm of bellmouth and taper, youll need to add some 50mm pipe to this to get the final length correct. that isnt the 'ideal' constant taper to the head, but thats life/fabrication limits!
Longer runners makes the acoustic 'ram effect' stronger. its still tuned to your specific peak RPM/power band, but it should also make more torque everywhere. the risk is that 'longer' (being greedy and wanting that stronger ram effect) can sometimes limit absolute peak flow due to friction (which causes sonic choke), but youre so far from choke i cant see the peak HP being effected in this case. theres no manipulation of trying to 'trick' your heads or cams, its simply free torque everywhere due to a well designed intake, essentially having your cake and eating it - thats why designing intakes is fun
I had one engine with a massive cam, that made power from about 7500-9500rpm. it wasnt even on cam yet when it went past the original motors rev limit (6500), and made more power at that rpm than the original motor at peak.
regarding the plenum... it must be possible to sneak it around the turbo, and make it a good shape.
if perhaps the 320mm runners are infact too long and make a packaging issue, you can go one harmonic shorter, and that would be around 260mm from valve to bell. could then potentially fit the plenum on top of the engine and not worry about hitting the turbo?"
This from the Intake person who recommended these lengths. These are from a series of emails. So they are somewhat informal but you will get the gist. In the end we had to use 1 size down from his suggested Velocity stacks. We're 48mm I.D. which have a 94mm Bell not the 50/100m he suggested.
"Ok, some numbers, ill just type them out in no particular order as I run the calcs
Firstly your port flow seems right on the money. I calculate max required intake port flow at 240-254cfm @28". Exhaust port at 203-220cfm @ 28"
Pipemax suggests you can keep pushing the lift, and it'll keep flowing, but we knew that already
Calculated power band 5200-7300rpm
Port velocities are going to be a bit slow based on your port size at flange (would need the exact area of the port to say exactly, but based on gross dimensions its on the bigger side of just acceptable) 3rd Harmonic tuning will prob work out best for you, taking packaging into consideration (this gives you longer runners and more torque, whilst still tuning for RPM specific peak HP)runner length (from back of valve) = 317mm looking at port speed and runner length, runner ID should be 50mm. That's a narrower cross section to your port at the flange, but a bit of forming and grinding will get a smooth enough transition.
Id be getting these: http://www.jntperformance.com.au/ind...l?catId=269035
Nice taper over the length, then just 50mm ID tube/bend to the flange what youll probably find is with 2bar of boost, youll lose a bit of that (effective) diamater to boundary layer effects, in essence itll act like its a bit smaller, and your velocities will pick up a bit (which in your case is a good thing). given the radius of those bellmouth, id make the length from bellmouth lip to valve 320mm."
"Intake acoustic harmonics always go in 'tone' steps. They're all designed to 'come on song' at the same particular frequency/note (ie rpm) but the energy in the harmonic goes down as you go to shorter and shorter steps (higher octave pitch). Longer gives more energy (low note), and make therefore more torque (everywhere), with the exception that as you go really long into really low harmonics, you start having major packaging problems (500mm long runners etc), and they have horrible losses at really high flow rates. OEM make really long, curly, spiral intakes which make great torque when the engine isn't flowing much at lower rpms, but can't handle the absolute flow demands for peak flow at high rpms.
You've got two length choices, one (from memory) around 250mm, and the one I gave you at 320mm. The longer one I think in your case, has no packaging disadvantage, and might actually help get your plenum in the right spot with the tb somewhere useful. It'll produce more tq, and I can't think that with a 50mm ID runner and a nice tapered bell, that you're going to have high flow issues. As I mentioned, if anything, it'll get the intake velocity a bit quicker, and you can use that (you're nowhere near choking velocity). If it were mine, I'd pick the longer runner.
Regarding plenum... A gentle bend to the runner is fine. I'd be guessing the ports in the head are about, I dunno, 80mm long (total guess). The bellmouths are 150mm long. That gives you ~90mm of space to slot in a bit of 50mm ID gentle mandrel bend. Bellmouths should be flat in the plenum base, you want the flow in the plenum body to be as symmetric as possible. Offset runners, angled bells etc... try and avoid
Of note, those bellmouths are HUGE... 100mm bell diameter, and 50mm ID at the end. It'll look nuts! For that size bell, I hope you've got enough room for that width of plenum. Ideally the plenum should be a bit wider, but fabrication could be a bit of a ****... 4" alloy tube and donuts and bends are easy to get. 5"? Not so much! Can stretch the 4" open a few mm to get it all to fit!
Length of your old manifold is kinda irrelevant, as the new cams, the new flow rates, and the bigger (50mm) runner dia determine the length, and have no recognizable relation to the old motor/manifold. And the port size at the flange in the head, don't lose sleep over it. Can't change it, and it's not 'bad' by any measure. Just on the bigger side. 50mm runner will match to it pretty neatly, and the velocity coming down the runner won't be too disturbed by a small step or change in port profile."
"320mm from bell to valve
the 150mm bellmouth from JNT is just 150mm of bellmouth and taper, youll need to add some 50mm pipe to this to get the final length correct. that isnt the 'ideal' constant taper to the head, but thats life/fabrication limits!
Longer runners makes the acoustic 'ram effect' stronger. its still tuned to your specific peak RPM/power band, but it should also make more torque everywhere. the risk is that 'longer' (being greedy and wanting that stronger ram effect) can sometimes limit absolute peak flow due to friction (which causes sonic choke), but youre so far from choke i cant see the peak HP being effected in this case. theres no manipulation of trying to 'trick' your heads or cams, its simply free torque everywhere due to a well designed intake, essentially having your cake and eating it - thats why designing intakes is fun
I had one engine with a massive cam, that made power from about 7500-9500rpm. it wasnt even on cam yet when it went past the original motors rev limit (6500), and made more power at that rpm than the original motor at peak.
regarding the plenum... it must be possible to sneak it around the turbo, and make it a good shape.
if perhaps the 320mm runners are infact too long and make a packaging issue, you can go one harmonic shorter, and that would be around 260mm from valve to bell. could then potentially fit the plenum on top of the engine and not worry about hitting the turbo?"
05-22-2015, 07:30 PM
#23
Three Wheelin'
I remember the Rimflow valves from decades ago, I think David Vizard came up with the design. Basically there was a groove machined into the cylinder face side of the inlet valve to hinder flow back into the inlet port, and a similar ridge on the back of the exhaust valve to prevent exhaust flow back into the cylinder. There's more to cylinder filling than port/manifold V valve size.
Thanks,
Mike
05-23-2015, 02:32 AM
#25
Three Wheelin'
...we return you to your usual programming...
05-23-2015, 02:42 AM
#26
Three Wheelin'
Hi Patrick,
Runners look good.
It's not my place to say, but I'm not a fan of the plenum design.
You might get a "piccolo" effect on the front cylinder, and a ram effect on the rear.
The piccolo effect isn't well documented, but it's like blowing across the top of a bottle - creates resonance and vaccuum. Had to redesign some industrial heat exchangers (made by Rolls Royce actually) that had this problem.
Time will tell. Are you going to run individual EGT?
Cheers,
Mike
Runners look good.
It's not my place to say, but I'm not a fan of the plenum design.
You might get a "piccolo" effect on the front cylinder, and a ram effect on the rear.
The piccolo effect isn't well documented, but it's like blowing across the top of a bottle - creates resonance and vaccuum. Had to redesign some industrial heat exchangers (made by Rolls Royce actually) that had this problem.
Time will tell. Are you going to run individual EGT?
Cheers,
Mike
05-23-2015, 04:15 AM
#27
Rennlist Member
Thread Starter
Mike
As you can imagine, there is a fair bit of 'inexact science' going on. One of the main issues we have is the packaging constraints from the cooling core duct. Having said that, we were trying to slow the flow upon entry as to not 'overlook' 1 plus make sure there was enough 'breathing space' around 4 as not to starve it. Tapered, log, centrally fed, ITB...then there's a reverse 4-1 header style that was recommended. We are definitely working within constraints, not the least me being project chief! EGT's are on the menu. Just waiting for the Specials Board.
Today's work from Paul.
As you can imagine, there is a fair bit of 'inexact science' going on. One of the main issues we have is the packaging constraints from the cooling core duct. Having said that, we were trying to slow the flow upon entry as to not 'overlook' 1 plus make sure there was enough 'breathing space' around 4 as not to starve it. Tapered, log, centrally fed, ITB...then there's a reverse 4-1 header style that was recommended. We are definitely working within constraints, not the least me being project chief! EGT's are on the menu. Just waiting for the Specials Board.
Today's work from Paul.
05-23-2015, 04:19 AM
#28
Rennlist Member
Thread Starter
Found the pic of the Header style inlet I mentioned. This is on a highly developed Valiant slant six nearby. Interesting concept.
05-23-2015, 09:44 AM
#29
Three Wheelin'
Hi Patrick,
Yep definitely a million and one parameters to juggle. Looking good so far! Your soap box derby car is shaping up nicely.
Epic effort, and I'm jealous.
Hope to make it across the ditch to see it some time.
Cheers,
Mike
P.s. The header style intake looks very cool but collectors don't work well in reverse. Aside from flow separation causing maldistribution, flow losses for an included angle of 45 degrees (22.5 degrees from centreline) are greater than for a step expansion.
Yep definitely a million and one parameters to juggle. Looking good so far! Your soap box derby car is shaping up nicely.
Epic effort, and I'm jealous.
Hope to make it across the ditch to see it some time.
Cheers,
Mike
P.s. The header style intake looks very cool but collectors don't work well in reverse. Aside from flow separation causing maldistribution, flow losses for an included angle of 45 degrees (22.5 degrees from centreline) are greater than for a step expansion.
05-23-2015, 11:32 AM
#30
Rennlist Member
Thread Starter
Mike it's so far beyond my paygrade that Time separates logic....however, each person that espouses their theory does it with absolute conviction so much so as to convince the working classes that there is no question as to their logic.