HIGHWAYMAN: Bringing the Devore 928 back from the dead
#571
Nordschleife Master
Everything should be checked, but it's my sense that the 39mm valves like those in the heads of this thread/build are not especially challenging in terms of intake/exhaust valve interference. Since this engine is to be run with a plenum manifold and since the porting doesn't appear to be geared towards killing low lift flow, I'd assume that the overlap chosen will be relatively conservative and thus there intake and exhaust valves should stay pretty far from each other. This is, of course, just guesses and conjectures on my part.
#572
Developer
I was afraid this would happen. ptuomov asked about valve overlap in post 546, and I answered there were only two adjustments possible: cam timing from cam relative to crank, and cam phasing relative intake to exhaust. These are things you guys already know. As we all agree - apart from something wild like that adjustable center sprocket cam that Strosek Ultra has shown, valve overlap is not really adjustable with what Porsche has given us to work with.
When I mentioned cam phasing in my answer to ptuomov, I did not mean to suggest we would be changing the cam phasing, actually, I rather doubt we will for the reasons already explained.
I agree. We have had cam phasing jump a tooth at the racetrack on our race motor, but it is only 8.5:1 CR so it kept running, was just down on power. This is a couple years ago now. But I have no plans to try that with Adam's motor. At 10.4:1 I don't think it wise
When I mentioned cam phasing in my answer to ptuomov, I did not mean to suggest we would be changing the cam phasing, actually, I rather doubt we will for the reasons already explained.
Also, there is every chance that you will have piston/valve contact, particularly on a higher comp/higher lift engine. Don't ask how I know, but I do.
#574
Nordschleife Master
#575
Developer
Extrude-Hone Results - pictures
So, we took the two intake manifolds, one dead-stock OEM and one after extrude-honing, to the flow bench. Both intakes have identical casting numbers. Same casting.
If you are coming into this conversation late, not sure of what extrude-honing is or would like to see some pics of the before-and-after, please scroll up to post 518 of this thread.
We tested each manifold two times, Intake runners for cylinder 3 and cylinder 5. Cylinder 5 was chosen because it was requested by ptuomov and because it is the one with the sharpest radius turn. Cylinder 3 was chosen as representative of one of the longer, milder runners.
The results show below. The colors are a little hard to read, but the chart at the bottom tells the story. Please ignore the column headers for valve lift as that has no bearings on this test.
TURBULENCE: We did take 4 measurements in 15 second intervals per runner, and if you read from L to R you will see small differences in the flow which is attributed to turbulence within the runner. This shows that the extrude-honed runners had smaller variables in flow over the 1-minute test, suggesting they have less turbulent and more consistent flow.
FLOW: The numbers show an increase of about 20 CFM on the short and twisty cylinder 5, and an increase of just about 50 CFM on the relatively straight and much longer cylinder 3.
If you are coming into this conversation late, not sure of what extrude-honing is or would like to see some pics of the before-and-after, please scroll up to post 518 of this thread.
We tested each manifold two times, Intake runners for cylinder 3 and cylinder 5. Cylinder 5 was chosen because it was requested by ptuomov and because it is the one with the sharpest radius turn. Cylinder 3 was chosen as representative of one of the longer, milder runners.
The results show below. The colors are a little hard to read, but the chart at the bottom tells the story. Please ignore the column headers for valve lift as that has no bearings on this test.
TURBULENCE: We did take 4 measurements in 15 second intervals per runner, and if you read from L to R you will see small differences in the flow which is attributed to turbulence within the runner. This shows that the extrude-honed runners had smaller variables in flow over the 1-minute test, suggesting they have less turbulent and more consistent flow.
FLOW: The numbers show an increase of about 20 CFM on the short and twisty cylinder 5, and an increase of just about 50 CFM on the relatively straight and much longer cylinder 3.
Last edited by Carl Fausett; 02-28-2017 at 11:17 AM.
#576
Developer
Here are a couple thoughts on the results...
The Widening Trend: Note that in the chart and the graph, the stock intake runners were actually quite close to each other in CFM. This even though one runner was short and hooked back on itself, and the other was much straighter and longer. After extrude-honing, the delta between those same two runners is now much wider.
IMO, this is a good thing. When designing or developing an intake manifold, it is common practice to use some longer runners and some shorter runners to create a nice long power band. If all the runners are identical, the engine output will be very peaky - not the best for a road racing application. I rather like that the trend between these two types of intake runners has moved apart from each other. I expect to see a wider and more useful power curve on the engine dyno as a result. This engine will be easier to drive.
The Match to the Heads Question: Was the intake manifold the restriction? Answer: Yes and No. Look at the flow-chart I posted on this thread at 509 (and I am posting it here again for your convenience). By comparing the stock head flow of one intake for one cylinder to the stock intake runner flow for one cylinder, you can see that the intake out-flowed the cylinder head by a comfortable little margin. (I am using .400" of lift as my comparable, you can compare any you like) They were nicely matched with each other, the intake flowing a little better than the head did.
Then we ported the heads and cut in larger intake valves. The flow tests for the ported heads now show that the heads out-flowed the intake manifold. (Again, I am using the CFM at .400" of lift) So, the intake manifold that wasn't the restriction had now become one. In order to get all the benefit out of those heads, the intake had to be opened up too, just as we suspected.
Finally, compare the numbers for the flow of the heads after porting to the intake runner after extrude-honing and the ratios are back where they should be - almost matched in flow, with the intake flowing a little better than the head.
This is why i am carrying our flow improvements out of the intake and even further to the intake shoe and the throttle body lest they become the next restriction.
The Widening Trend: Note that in the chart and the graph, the stock intake runners were actually quite close to each other in CFM. This even though one runner was short and hooked back on itself, and the other was much straighter and longer. After extrude-honing, the delta between those same two runners is now much wider.
IMO, this is a good thing. When designing or developing an intake manifold, it is common practice to use some longer runners and some shorter runners to create a nice long power band. If all the runners are identical, the engine output will be very peaky - not the best for a road racing application. I rather like that the trend between these two types of intake runners has moved apart from each other. I expect to see a wider and more useful power curve on the engine dyno as a result. This engine will be easier to drive.
The Match to the Heads Question: Was the intake manifold the restriction? Answer: Yes and No. Look at the flow-chart I posted on this thread at 509 (and I am posting it here again for your convenience). By comparing the stock head flow of one intake for one cylinder to the stock intake runner flow for one cylinder, you can see that the intake out-flowed the cylinder head by a comfortable little margin. (I am using .400" of lift as my comparable, you can compare any you like) They were nicely matched with each other, the intake flowing a little better than the head did.
Then we ported the heads and cut in larger intake valves. The flow tests for the ported heads now show that the heads out-flowed the intake manifold. (Again, I am using the CFM at .400" of lift) So, the intake manifold that wasn't the restriction had now become one. In order to get all the benefit out of those heads, the intake had to be opened up too, just as we suspected.
Finally, compare the numbers for the flow of the heads after porting to the intake runner after extrude-honing and the ratios are back where they should be - almost matched in flow, with the intake flowing a little better than the head.
This is why i am carrying our flow improvements out of the intake and even further to the intake shoe and the throttle body lest they become the next restriction.
Last edited by Carl Fausett; 02-27-2017 at 06:56 PM.
#578
Banned
Thread Starter
Folks,
I've been in Stuttgart for the past couple of days; it was an impromptu visit because I had an opportunity to do a factory tour that I din't want to pass up.
Being here is surreal yet a little dissapointing at the same time. For example, I walked past Werk 1 at the factory where my 356 would've been assembled in 1962, and probably where some of my other cars were assembled as well. This is thoroughly overshadowed by the larger modern assembly line next door, where Porsche pumps out a new car (718, 911) every 3.5 minutes. Yes, you heard that right, they build on average 224 boxsters, Caymans, and 911s here every day.
The disappointment comes from the lack of knowledge or caring about the older models here at the factory; the only people that really care about the past can be found in the Porsche museum across the street.
It's a little sad because our cars (even the Devore car) were mostly assembeled by hand. Engines were build by hand. Mind you, there are still no robots on the main chassis assembly line here today, but you get the impression that the workers are less skilled than their past counterparts.
I watched as a flat 6 was moved to a new station on a coveyor. The motor stopped just long enough for a worker to position an oil pan on the bottom of the motor, then it moved to the next station, where a robot screwed in the oil pan bolts. This guy was doing nothing all day except mounting oil pans.
Down the line some, another guy was torquing head studs. He had a torque wrench mounted to a cord from the cieling. It had no numbers of any kind on it; it would just flash a yellow light when the torque value was achieved. I almost wonder if the guy even knew what the correct value was, or if he was just turning the wrench until the light came on.
Many of us know more about building engines than the factory workers probably do, LOL.
Pics later when I can upload them, although none will be of the factory - they take your phones and cameras at the door.....
I've been in Stuttgart for the past couple of days; it was an impromptu visit because I had an opportunity to do a factory tour that I din't want to pass up.
Being here is surreal yet a little dissapointing at the same time. For example, I walked past Werk 1 at the factory where my 356 would've been assembled in 1962, and probably where some of my other cars were assembled as well. This is thoroughly overshadowed by the larger modern assembly line next door, where Porsche pumps out a new car (718, 911) every 3.5 minutes. Yes, you heard that right, they build on average 224 boxsters, Caymans, and 911s here every day.
The disappointment comes from the lack of knowledge or caring about the older models here at the factory; the only people that really care about the past can be found in the Porsche museum across the street.
It's a little sad because our cars (even the Devore car) were mostly assembeled by hand. Engines were build by hand. Mind you, there are still no robots on the main chassis assembly line here today, but you get the impression that the workers are less skilled than their past counterparts.
I watched as a flat 6 was moved to a new station on a coveyor. The motor stopped just long enough for a worker to position an oil pan on the bottom of the motor, then it moved to the next station, where a robot screwed in the oil pan bolts. This guy was doing nothing all day except mounting oil pans.
Down the line some, another guy was torquing head studs. He had a torque wrench mounted to a cord from the cieling. It had no numbers of any kind on it; it would just flash a yellow light when the torque value was achieved. I almost wonder if the guy even knew what the correct value was, or if he was just turning the wrench until the light came on.
Many of us know more about building engines than the factory workers probably do, LOL.
Pics later when I can upload them, although none will be of the factory - they take your phones and cameras at the door.....
#579
Rennlist Member
Join Date: Feb 2011
Location: Mostly in my workshop located in Sweden.
Posts: 2,241
Received 475 Likes
on
251 Posts
I was afraid this would happen. ptuomov asked about valve overlap in post 546, and I answered there were only two adjustments possible: cam timing from cam relative to crank, and cam phasing relative intake to exhaust. These are things you guys already know. As we all agree - apart from something wild like that adjustable center sprocket cam that Strosek Ultra has shown, valve overlap is not really adjustable with what Porsche has given us to work with.
When I mentioned cam phasing in my answer to ptuomov, I did not mean to suggest we would be changing the cam phasing, actually, I rather doubt we will for the reasons already explained.
I agree. We have had cam phasing jump a tooth at the racetrack on our race motor, but it is only 8.5:1 CR so it kept running, was just down on power. This is a couple years ago now. But I have no plans to try that with Adam's motor. At 10.4:1 I don't think it wise
When I mentioned cam phasing in my answer to ptuomov, I did not mean to suggest we would be changing the cam phasing, actually, I rather doubt we will for the reasons already explained.
I agree. We have had cam phasing jump a tooth at the racetrack on our race motor, but it is only 8.5:1 CR so it kept running, was just down on power. This is a couple years ago now. But I have no plans to try that with Adam's motor. At 10.4:1 I don't think it wise
Åke
#580
Rennlist Member
Join Date: Feb 2011
Location: Mostly in my workshop located in Sweden.
Posts: 2,241
Received 475 Likes
on
251 Posts
Folks,
I've been in Stuttgart for the past couple of days; it was an impromptu visit because I had an opportunity to do a factory tour that I din't want to pass up.
Being here is surreal yet a little dissapointing at the same time. For example, I walked past Werk 1 at the factory where my 356 would've been assembled in 1962, and probably where some of my other cars were assembled as well. This is thoroughly overshadowed by the larger modern assembly line next door, where Porsche pumps out a new car (718, 911) every 3.5 minutes. Yes, you heard that right, they build on average 224 boxsters, Caymans, and 911s here every day.
The disappointment comes from the lack of knowledge or caring about the older models here at the factory; the only people that really care about the past can be found in the Porsche museum across the street.
It's a little sad because our cars (even the Devore car) were mostly assembeled by hand. Engines were build by hand. Mind you, there are still no robots on the main chassis assembly line here today, but you get the impression that the workers are less skilled than their past counterparts.
I watched as a flat 6 was moved to a new station on a coveyor. The motor stopped just long enough for a worker to position an oil pan on the bottom of the motor, then it moved to the next station, where a robot screwed in the oil pan bolts. This guy was doing nothing all day except mounting oil pans.
Down the line some, another guy was torquing head studs. He had a torque wrench mounted to a cord from the cieling. It had no numbers of any kind on it; it would just flash a yellow light when the torque value was achieved. I almost wonder if the guy even knew what the correct value was, or if he was just turning the wrench until the light came on.
Many of us know more about building engines than the factory workers probably do, LOL.
Pics later when I can upload them, although none will be of the factory - they take your phones and cameras at the door.....
I've been in Stuttgart for the past couple of days; it was an impromptu visit because I had an opportunity to do a factory tour that I din't want to pass up.
Being here is surreal yet a little dissapointing at the same time. For example, I walked past Werk 1 at the factory where my 356 would've been assembled in 1962, and probably where some of my other cars were assembled as well. This is thoroughly overshadowed by the larger modern assembly line next door, where Porsche pumps out a new car (718, 911) every 3.5 minutes. Yes, you heard that right, they build on average 224 boxsters, Caymans, and 911s here every day.
The disappointment comes from the lack of knowledge or caring about the older models here at the factory; the only people that really care about the past can be found in the Porsche museum across the street.
It's a little sad because our cars (even the Devore car) were mostly assembeled by hand. Engines were build by hand. Mind you, there are still no robots on the main chassis assembly line here today, but you get the impression that the workers are less skilled than their past counterparts.
I watched as a flat 6 was moved to a new station on a coveyor. The motor stopped just long enough for a worker to position an oil pan on the bottom of the motor, then it moved to the next station, where a robot screwed in the oil pan bolts. This guy was doing nothing all day except mounting oil pans.
Down the line some, another guy was torquing head studs. He had a torque wrench mounted to a cord from the cieling. It had no numbers of any kind on it; it would just flash a yellow light when the torque value was achieved. I almost wonder if the guy even knew what the correct value was, or if he was just turning the wrench until the light came on.
Many of us know more about building engines than the factory workers probably do, LOL.
Pics later when I can upload them, although none will be of the factory - they take your phones and cameras at the door.....
Åke
#581
Nordschleife Master
Runner diameter?
How much did the runner cross sectional area increase with extrude honing? I'm guessing about 7.5% or so increase in diameter?
The reason why I am asking is that I'd find it most interesting to know what cross-sectional area extrude hone resulted in (+1mm or +2mm increase in diameter, or what) and how the CFM/square inch of area compares between stock and modified runners.
The reason why I am asking is that I'd find it most interesting to know what cross-sectional area extrude hone resulted in (+1mm or +2mm increase in diameter, or what) and how the CFM/square inch of area compares between stock and modified runners.
Last edited by ptuomov; 02-28-2017 at 09:19 AM.
#583
Developer
I'm not sure the runner diameters matter as much as the smoothness that has been gained. It is a rough casting, and the roughness creates a larger boundary layer and in places may even cause boundary layer separation. Both adversely effect the actual through-put.
Because of this, I feel the flow-bench data is more significant than a diameter measurement.
That said, here is what I measured. Keep in mind that the runners are not perfectly round, and that they are a casting. I tried to pull reliable measurements from the same spot on the two manifolds at the same angle. On average, the OEM manifold had diameters between 1.760" and 1.769". On average, the extrude-honed intake runners had diameters between 1.775" and 1.778". They were also more consistent, which I attribute to the reduction in surface roughness.
Because of this, I feel the flow-bench data is more significant than a diameter measurement.
That said, here is what I measured. Keep in mind that the runners are not perfectly round, and that they are a casting. I tried to pull reliable measurements from the same spot on the two manifolds at the same angle. On average, the OEM manifold had diameters between 1.760" and 1.769". On average, the extrude-honed intake runners had diameters between 1.775" and 1.778". They were also more consistent, which I attribute to the reduction in surface roughness.
#584
Nordschleife Master
I'm not sure the runner diameters matter as much as the smoothness that has been gained. It is a rough casting, and the roughness creates a larger boundary layer and in places may even cause boundary layer separation. Both adversely effect the actual through-put.
Because of this, I feel the flow-bench data is more significant than a diameter measurement.
That said, here is what I measured. Keep in mind that the runners are not perfectly round, and that they are a casting. I tried to pull reliable measurements from the same spot on the two manifolds at the same angle. On average, the OEM manifold had diameters between 1.760" and 1.769". On average, the extrude-honed intake runners had diameters between 1.775" and 1.778". They were also more consistent, which I attribute to the reduction in surface roughness.
Because of this, I feel the flow-bench data is more significant than a diameter measurement.
That said, here is what I measured. Keep in mind that the runners are not perfectly round, and that they are a casting. I tried to pull reliable measurements from the same spot on the two manifolds at the same angle. On average, the OEM manifold had diameters between 1.760" and 1.769". On average, the extrude-honed intake runners had diameters between 1.775" and 1.778". They were also more consistent, which I attribute to the reduction in surface roughness.
First, there's the cross-sectional area. That area needs to be changed either if the original designer of the engine was incompetent (unlikely in this case) or had different design objects or if the rest of the engine has changed (for example to a larger displacement).
Second, there's the flow per cross-sectional are, basically CFM/sqin. The CFM/sqin is often normalized by the CFM/sqin from an ideal orifice, the result of the computation being Cd or coefficient of discharge. Unlike with the cross-sectional area, which has an optimum which is better than either smaller or larger, higher Cd is basically better.
If your measurements are accurate, then I'd say the Cd has improved for the extrude honed manifold.
#585
Developer
I am used to seeing Cd as Coefficient of Drag, not Discharge. But if I understand you correctly, you are seeing higher flow greater than what would be gotten by the change in diameter alone, which suggests turbulance in the runner has been reduced, the boundry layer is now thinner, freeing up more of the tube for laminar flow.
That is our conclusion, and I think what you are saying as well. Let me know if you agree.
That is our conclusion, and I think what you are saying as well. Let me know if you agree.