Why people should heed the advice.....
Newbie
Joined: Jun 2004
Posts: 30
Likes: 0
From: Tokyo, Japan
First I want to say that I've talked to Indy EVO over PM and the waters are now calm and we'll probably even be sharing info in 4g63 land speed racing efforts for 08 season its going to be an awesome year.
Holy #$&% theres a lot to reply to, so very long post will follow:
It is a really small world some times. A long time friend of mine's father is Tony Manzer from Comptech their engine builder. You might know him? Hes the real deal. One of those guys who doesn't say much but I'll spend a long time still learning from one sentence. Very quiet, pure genious. What happened and is happening with those cars on those years was really something else. Thats not an easy series to win. Kudos to you.
Mitsubishi stock ECU fuel trims are funky I know even idling for 10 minutes will change. I know the basics of them but I know when I don't know enough to say. I do know who would know. Pete at Works, he knows that stock ECU inside and out so I'll try to get a hold of him and ask but hes tough to get a hold of it might be a while till I hear back. Probably now that we are actually having a useful discussion here it will be easier to enlist help so I'll ask.
It could be more complex than that I think its good we're going to design some testing and ask the ECU gurus
Yah, theres a couple ways I had in mind. The first thing is that you have to go to knock threshold to know where it is so you have to use a gas where that threshold isnt higher than anything you can throw at it (like C16). Since DTM has an AEM he can log the knock volts channel which is filtered knock sensor for after ignition events. If his geometry is close to stock the stock knock sensor does an excellent job they have already tuned it to the specific frequencies that our engines produce knock at. Failing that I would suggest starting at a very low boost level and tuning to the point where you loose power either advancing or retarding timing then you see if that threshold moves on manifold A to manifold B this can be tested without a dyno too.
Any thoughts on this? That seems like a lot of work in comparison to just logging knock data on the AEM. I asked if he has a motec becuase spare inputs on motec are hard to come by and I could lend some very high frequency data aquisition hardware to the test rig at least for a couple weeks till my car is back together.
If it was my car I would test on pump gas, turn up the boost till I saw knock then back it off. Play with timing for the same thing. A tune can make power in lots of ways. It's possible one part makes your car make power with more timing than boost and another with different fueling. I've never tuned two cars that were the same. In general theres a 4g63 theme but when you get down to throwing 18hp around, I feel that there are a lot of intricacies.
I'll tell you the truth. When I start trying to picture the stuff going on in manifolds my head spins. Those things you listed interact in ways that I can understand but just cant calculate all at once. Theres all the stuff that happens with the runners are curved, tapered, diameter, boost pressure, thats why I like to go and fluid model parts like that. Its just too much to keep track of in my head. Same with multi link suspensions I stare at the animation for hours and then just ask the computer what its doing. I'm sure there are people who can do it but I shouldnt answer because my answer would be "lets build a model and test it all". If you have some patience I cant talk about it but we are not the first to ask these questions and someone is a step ahead on that at least. So I'd rather wait for that to go public and lean on the data than go through it all on my own. Do you do any CFD for your manifolds? Or are you one of those guys better than myself who can do it all in their heads?
Your point here is well taken, and while i have not done the legwork to determine what kind of heat effects are driving which results, I KNOW they have an effect. This is why I was very interested to see all the runs that David had done. My own car will vary 20hp from run to run on heat soak alone. Exactly what thing is heat soaking and why its causing the effect I do not know and I want to investigate but I have had so many things higher on the list than 20hp dissapearing on the dyno. But I've been slowly measuring temperature deltas on critical systems and trying methods of cooling them. Some I can combat and others I cannot. I started with most critical systems like intercooler and fuel. I now gain 3 degrees air temperature across an entire 0-180pass in 110 degrees ambient I am moving through one system at a time.
What I do in order to avoid this effect on the dyno is I test parts until heat soak has stabilized. Which is 3-4 consecutive dyno runs until the car is down about 20hp then the deviations go down to about 2-3 HP so I call that my standard deviation. Any part that nets less then 2-3HP I cannot tell if it makes more power becuase thats my standard deviation. Which kind of goes to show how you could test any part to loose power you just compare a certain run to a certain run. Thats why its critical to perform these kind of tests without an agenda.
There are actually a lot of things your talking about in that post, I'm strugging to keep up I am trying to get some work done and slowly making my way through this reply. Yes I agree intake valve is the largest restriction, I would rank the throttle plate second though. When I did some data modeling I found the air passing through a 65mm throttle body was moving at an incredible rate and the turbulence from the throttle plate was effecting air density 2-3% all the way down to the valves. Thats why I went to the Q45TB and 3.5" piping I netted 28hp there.
You are right though there are SO many complexities that we can only pick a few theories to test on. It's difficult to decide and thats why going faster is never so fully defined and is an iterative process. We also don't have available to us the high end test equipment that would make these things easier. I think your on the level of building getting 27 more hp out of an intake manifold that was already badass and we are on the base level of trying to get one that works at all. As much as I want to address all those intricacies because I think they are truly where the car gets to be awesome I think this development cycle is at a much lower level. Would you agree?
Ahh yes the AEM. I actually have the first one ever made and oh how I love and hate it. I used an autronic on my old bike and loved the thing. Can you still even get those? I heard autronic was gone now? The new MoTeC systems will be badass. If I didnt have like 100 hours into this AEM tune I would have switched already.
Thats true the water temps on the other side of the head can be drastic changes. But I wouldnt account for 40hp that way, would you? I heat soak the **** out of my car in every way on the dyno and I've never seen a 40hp swing. Have you?
agreed and its all speculation on which one goes which way but I think from this would could compose a laundry list of things that could be tested for a version 2 of intake manifold testing. I just dont want to see us end up building a list of like 10,000 variables for poor DTM to test and he doesnt even get his cars ready in time for season. My suggestion is to pick a few front runner theories and test em.
I disagree. Even though many many people have their ways of conducting their procedures...this is something I keep as a constant. Depending on HOW you are circulating airflow to mimic real world loading conditions, this should not be an UNCONTROLLABLE problem. What I mean is you should have control over what range the car is being tuned to at all times. I prefer to test the vehicle under the worst possible condition. It allows us to see many extremes. Whether its allowing a specific amount of time to cool between runs, or constant loading and hard pulls back to back, we log our results and tune the vehicle in a manner that will be as consistent as possible, while maintaining power and reliability.
Everybody has their own way.
Maybe Indy is thinking about the kind of racing we do too, imagine if you race land speed. For us there no way to emulate our conditions. 110+ degrees ambient at 5000ft altitude with 210mph dust filled air blowing over your car. Youd need like a wind tunnel + dyno from hell. It would probably cost a million dollars for a single test. Maybe thats what hes thinking of but your right too, we can only test with what limited means we have and do the best we can.
It seemed like Marco was a lot more forthcoming with his data for people to speculate and tear it apart. Half of the first questions I had in my mind about a test like this were laid to rest either in the data logs or videos of Marcos testing. I just couldnt say the same for David's, which is why I was questioning him. I wanted answers and I still haven't gotten them. With Marcos we are doing what we are doing now, pouring over the data and trying to understand. As usual with any test we walk away with more questions than answers and somehow that is satisfying, all I know right now is its saturday and I need a beer, real bad.
Holy #$&% theres a lot to reply to, so very long post will follow:
2002 w/ Robby Gordon I also spotted for him in the race. The intake was sold to Comptech who had it on Jeff Wards car the year after for Ganassi. I know two engine builders GW, GH, one tuner Billy, and one engineer in England. That manifold was worth 27 HP. By the way were working on the same project the one with closed wheels. You know thats about all I can say about it. Talk to GW and see what he thinks of my resume. C Ya
I'm just looking for a good explanation for what the data log shows. To me, it really looks like a timing advance difference being the source of the power gains. But I'm all ears on any good explanation behind what is seen in the logs.
Gixer, I'm interested in how you want to test it. Care to elaborate?
Any thoughts on this? That seems like a lot of work in comparison to just logging knock data on the AEM. I asked if he has a motec becuase spare inputs on motec are hard to come by and I could lend some very high frequency data aquisition hardware to the test rig at least for a couple weeks till my car is back together.
If it was my car I would test on pump gas, turn up the boost till I saw knock then back it off. Play with timing for the same thing. A tune can make power in lots of ways. It's possible one part makes your car make power with more timing than boost and another with different fueling. I've never tuned two cars that were the same. In general theres a 4g63 theme but when you get down to throwing 18hp around, I feel that there are a lot of intricacies.
Gixxer Drew, I have a question without ego or what has been said in the past posts. Looking at the from the outside all things being equal, not looking at this as a Mangus or other manifold thing. With the cylinder head and camshafts being a constant for the test. A manifold consists of a plenum and runners with a TB on the end of it with the inlet size to the cylinder head fixed to a constant area and the turbo pressure and TB being equal for both tests is this correct?
One would think the only changes possible are the plenum volume, runner taper, and radius for the runners from the plenum.
One would think the only changes possible are the plenum volume, runner taper, and radius for the runners from the plenum.
This might be true, but at equal boost levels across the testing range in a controled enviroment ie. engine dyno with controled cooling. Can you explain why the logs ie. fuel curve,timing, and water temp shows a change? Other than VE and fill changes at WOT the major restriction is the intake valve would you agree? but with constant pressure on the backside of the valve before the valve opens and given the amount of max airflow the cylinder head can move as constant, the manifold turns into a storage container. All things being equal here the runner inlet may be the key but, here is another question to ponder while the car ran for 3 hours cooled for a couple what do you think about heat soak being a factor in the first runs as the casting will not transfer heat out as quickly as a sheetmetal part, also the injectors as they are an electronic part will heat soak quickly without fuel running thru them, as you know that heat will affectly will change snap rate ie. opening and closing rate which I have seen in the past. What about the transfer of under car heat to the fuel lines from the point nearest to any heat source back to the tank with the initial runs could that possibly have skewed results untill the fuel temp has reached a constant? With the manifold change over how much time elapsed? Could this also contribute to a lower fuel and coolant temp? Also could the initial runs data be a result from heat soak since as it was stated the hood was not lifted until it was run on the dyno? Just a few qustions to ponder while trying to make sense of why did results vs the trims not jive. No finger pointing just a discussion.
What I do in order to avoid this effect on the dyno is I test parts until heat soak has stabilized. Which is 3-4 consecutive dyno runs until the car is down about 20hp then the deviations go down to about 2-3 HP so I call that my standard deviation. Any part that nets less then 2-3HP I cannot tell if it makes more power becuase thats my standard deviation. Which kind of goes to show how you could test any part to loose power you just compare a certain run to a certain run. Thats why its critical to perform these kind of tests without an agenda.
There are actually a lot of things your talking about in that post, I'm strugging to keep up I am trying to get some work done and slowly making my way through this reply. Yes I agree intake valve is the largest restriction, I would rank the throttle plate second though. When I did some data modeling I found the air passing through a 65mm throttle body was moving at an incredible rate and the turbulence from the throttle plate was effecting air density 2-3% all the way down to the valves. Thats why I went to the Q45TB and 3.5" piping I netted 28hp there.
You are right though there are SO many complexities that we can only pick a few theories to test on. It's difficult to decide and thats why going faster is never so fully defined and is an iterative process. We also don't have available to us the high end test equipment that would make these things easier. I think your on the level of building getting 27 more hp out of an intake manifold that was already badass and we are on the base level of trying to get one that works at all. As much as I want to address all those intricacies because I think they are truly where the car gets to be awesome I think this development cycle is at a much lower level. Would you agree?
Cant really afford MoTec. lol Fuel Injection Angle and position in relation to the head port.
Again just for discussion purposes, you may have answered it below. Dyno operation /operator is critical to consistent and accurate results. Was the log showing calculated load? Do we have a copy of the original rom? (if the stock ecu is used) How do we know what load site the computer used in order to lead the engine? Water temp in the housing may not be as accurate as we may want it to be.
Again just for discussion purposes, you may have answered it below. Dyno operation /operator is critical to consistent and accurate results. Was the log showing calculated load? Do we have a copy of the original rom? (if the stock ecu is used) How do we know what load site the computer used in order to lead the engine? Water temp in the housing may not be as accurate as we may want it to be.
Alot of these points literally can go both ways, and may not give an advantage to either component in my opinion.
I disagree. Even though many many people have their ways of conducting their procedures...this is something I keep as a constant. Depending on HOW you are circulating airflow to mimic real world loading conditions, this should not be an UNCONTROLLABLE problem. What I mean is you should have control over what range the car is being tuned to at all times. I prefer to test the vehicle under the worst possible condition. It allows us to see many extremes. Whether its allowing a specific amount of time to cool between runs, or constant loading and hard pulls back to back, we log our results and tune the vehicle in a manner that will be as consistent as possible, while maintaining power and reliability.
Everybody has their own way.
Another thought could it be injector angle? I have not seen both manifolds side by side, but I know for a fact that this change can make HP. The manifolds are limited in design by the area it has to fit into also, the location of the TB is dictated by the same area or package, therefore most manifolds are very close in design. The bellmouth at the runner looks to be a Kinsler part grafted into the plenum this in it's self helps at the transition from N/A to the pont when the PSI starts building great for street driving after that point PSI is PSI only so much air can pass through a given orifice. So this in turn points to a volume issue, can the volume effect the response? yes it can and yes it does. Just as it has been stated we can't go with Buschur's results because you dont agree with his testing methods this would have to be the same for Marco. I have never personaly met Dave (spoke with him on the phone everybody knows how busy he is with the X stuff) or Marco or seen their testing facilities, but discounting one for another just because you happen to be a customer of one or the other would be not giving this evaluation a fair shake. One thing that I have seen with chassis dynos, they have good setups to cool the rad and the intercooler but nothing under the car, and engine compartment to keep the air being sucked in at a constant air temp or clean from exhaust contamination. This alone can skew the readings. It would be a positive to use a remote air source on the chassis dyno like the engine dynos use the reason being air in the dyno cell air becomes contaminated even with extractors on the exhaust and in the room. This in turn it very possible even with a open door and the fans on the engine to still contaminate the area with unburnt fuel and exhaust. Just a few more things to think about.
Last edited by gixxer_drew; Feb 23, 2008 at 01:49 PM.
...Yes I agree intake valve is the largest restriction, I would rank the throttle plate second though. When I did some data modeling I found the air passing through a 65mm throttle body was moving at an incredible rate and the turbulence from the throttle plate was effecting air density 2-3% all the way down to the valves. Thats why I went to the Q45TB and 3.5" piping I netted 28hp there...
l8r)
Evolved Member
Joined: Jul 2006
Posts: 542
Likes: 0
From: wishing i still lived in Spokane
wow i just finished reading the last 4 pages ive been watching this thread since it started...some good discussion going on in regards to the testing of "said' manifold and the dyno time, i hope all this gets sorted out....soon
Newbie
Joined: Jun 2004
Posts: 30
Likes: 0
From: Tokyo, Japan
Could you elaborate on that statement? At what power levels are you making those kinds of gains? I run an AMS VSR intake manifold with a 65mm TB, what do you think the potential gains - if any - would be to go to a 75mm TB on a pump gas tune (currently at about 400whp)?
l8r)
l8r)
I made those gains on 91 octane pump gas I just rechecked the dyno sheets and I have to make a slight correction, it was 22hp actually went from 443 to 465whp at 18psi back then I was on a 60trim but I now have an HTA35R i have yet to test. That was a stock DSM 60mm throttle body to a Q45. How that would apply to your setup there are way too many variables to say what kind of gains you could get your manifold is very different as is your intake ports compared to a 1g dsm. I also modified my throttle plate for aerodynamics I filled with epoxy and shaped to teardrop and machined a spacer to move the throttle plate even further from the plenum. I do know this, 2.5" piping the air was exceeding 200mph at the power levels I was at. In fluid modeling I had a 2% increase in air density at the end of the runners and then the real world result exceeded the model in terms of power gains by more than double, my only possible explanation was that the huge intake pipe was acting like a plenum. Hope that helps but as usual more questions than answers.
http://gallery.mac.com/andrewbrillia...&bgcolor=black
also I tried to get as much straight pipe as possible to have the air as smooth as possible before getting to the throttle plate so I moved the battery elsewhere and put the piping right up against the strut tower this also may have contributed. The 2.5" piping still had turbulence to the throttle plate from the same bends whereas the 3.5" did not due to the lower velocity. Before that I had short route piping which according to my data was actually hurting power becuase of the big bend right before the throttle plate. People said the huge piping was going to be more laggy, it was not. It also made the tial BOV feel like a stock BOV. Its like a little puppy dog. I liked the no boost leaks and prefect throttle response I was very very happy with that mod
Last edited by gixxer_drew; Feb 23, 2008 at 04:27 PM.
Thread Starter
Account Disabled
iTrader: (4)
Joined: Jun 2007
Posts: 754
Likes: 0
From: Dulles, VA 20166
Thats true the water temps on the other side of the head can be drastic changes. But I wouldnt account for 40hp that way, would you? I heat soak the **** out of my car in every way on the dyno and I've never seen a 40hp swing. Have you?
No honestly I cannot say that is the case. Just A variable in the equation of power potential. If I really recall I have seen no more than 10-12 WHP and about 4-6 ftlbs of torque vary from logged coolant temp changes.
Again though I have not SEEN the data for the 40 hp discrepancy. Just throwing our some ideas.
I'm actually excited over the testing. It should put some of this to rest.
I've seen multiple intake manifolds at DTM all huddled around 2-3 turbos. But DTM is being hush hush about what is going on with them. I think it is for this test and their twin scrolls...
Thread
Thread Starter
Forum
Replies
Last Post
