A day that turned into a week, on the flow bench.

davidbuschur · Feb 14, 2009, 05:41 PM

On the BR forums I have been conducting testing this entire last week. My goal was to figure out what makes an intake manifold work, how to improve ours and eventually head to the dyno for testing. What ended up happening this week was very un-expected. The entire week's work of posts I am going to combine here in this one post for anyone interested.
------------------------------------------------------------------------------------------------------
2-10-09

Today I spent the majority of the day on the flow bench. The findings were quite interesting.

The numbers below are the cylinder and CFM. All testing was done at 28" on a JKM Flow bench by BRS head services.

Here are the results of what I have found so far.

EVO8 Stock intake, w/stock throttle body:
#4-327, #3-308 (-6%), #2-300 (-9%), #1-300 (-9%), total 1235 CFM

BR Ported intake, w/65mm throttle body:
#4-336, #3-319 (-5%), #2-306 (-9%), #1-300 (-12%), total 1261 CFM

BR ported intake, w/65mm throttle body, additional port work
#4-337, #3-318 (-5%), #2-310 (-8%), #1-300 (-12%), total 1265 CFM

BR ported intake, w/65mm throttle body, even more port work
#4-331, #3-319 (-3%), #2-315 (-5%), #1-300 (-10%), total 1264 CFM

Wilson V2 intake, w/65mm throttle body
#4-330 (-3%), #3-329 (-3%), #2-321 (-5%), #1-340, total 1320 CFM

Driven Innovations, w/65mm throttle body
#4-320 (-5%), #3-324 (-4%), #2-334 (-1%), #1-337, total 1315 CFM

Then we tested the stock intake manifold with no port work done to it. The plenum cut off basically flush to the top of the runners. We put that on the flow bench and checked it.

#1-308cfm(-12%), #2-320cfm(-7%), #3-341cfm(-1%), #4-345cfm

This was quite shocking as the #1 runner hardly gained anything at all. The plenum/runners didn't seem to be the biggest problem in the intake at this point.

We then did some additional port work on just #1 and it changed nothing, still flowed 308 cfm.

Then we started playing with clay in the port to try to pick it up, again, no major changes. There was one last shot at making it flow better with something we are capable of doing here at the shop so I had Trent weld up part of the #1 port and did some grinding on it.

That port then went up to 316cfm with no plenum on the intake manifold.

I came back to the shop and grabbed the stock BR ported intake we were working with yesterday and duplicated what we did on the cut apart intake, the best I could. It's very hard to reach that runner in the intake as it's all the way in the back of the plenum.

The next flow test was very good and is as good as we can get it without trying to cut apart an intake manifold and copy someone's else's work, which I have no desire to do.

So here's how we ended up on the final port work on the stock intake:

#1-314cfm(-5%), #2-308 cfm(-7%), #3-315cfm(-4%), #4-330cfm
Total CFM 1267.

I put the i/c pipe on the throttle body at the angle it is installed on the car. The CFM's dropped about 5 CFM, so it lost some flow. I then turned it as far in one direction as I could and then rotated it back the other direction as far as I could. So over 180 degrees in rotation. By doing this the most it would effect any runner was less than 2 CFM of loss or gain. Very small change, seems like the biggest restriction is the pipe alone and that moving it around has little/no effect on air flow from there. Our pipe does have a 90 degree bend in it but remember after the bend it has about 2" straight and the length of the throttle body/intake manifold flange is another 3.5", I believe that length of straight is getting the air flow stabilized again and moving straight into the throttle body.

Interesting. I figured out the cubic inches of the plenum on the stock intake today. Measuring the plenum and stopping at where the runners form the plenum is only 52.32 cu. in. Very small considering the power the car makes on it.

I keep reading these theories on the plenum needing to be atleast 150% larger than the engine displament. That would mean the plenum would need to be at least 183 cubic inches.

I personally don't buy it from what I have seen. I think it adds lag and don't think it needs to be nearly that big. I guess looking at the stock plenum being 52 cu. in. kind of verifies that thought, to a point.

Runners. The stock runners in #4 measure close to 10", the V2 on my car measured 6.25" on runner #4, the DI measures 6.5" and the one we ported on measured 8.5".

2-13-09

I then put my RS on the dyno and dyno'd the V2 that was on my car and then the ported stock intake we did the extra work to.

While the two tests above were being conducted Ted was porting another stock intake manifold for me with our basic port work done to it. The flow data can be found at the top of this post.

This has been a hell of a day.

The day started off with putting my car on the dyno. For the comparison testing I was not going to run the car on kill so I just flipped the dual stage boost switch to low, which is 30 psi.

The car is set up exactly like it was a few weeks ago when it made 705 whp on the Wilson V2. I was running about 42 psi of boost then.

I put the car on the dyno and ran it to get all the fluids up to tempurature.

The first pull showed 607 whp/460 ft lbs of torque. I thought that was pretty strong. I did two back to back pulls, power on both were identical.

I then pulled the V2 off and put the BR ported stock intake on, this is the one we have been working on all week on the flow bench. I did two pulls again and the car made 610 whp and 468 ft lbs. To say I was completely unglued with happiness would be an understatement.

The base pull and all pulls were run to 8,000 rpm. IF I had thought I'd have seen what I ended up seeing on the graphs I have run the car further in the RPM. On the comparison you can see the Wilson V2 is still climbing at 8,000 rpm while the other runs are tapering off. This is important to note because for a really high rpm application the V2 looks like it will make more power over about 8300 rpm than what else I tested today.

OK, so now I wanted to test our standard ported intake manifold, which through all of this crap I have never tested on the dyno against other intake manifolds. There wasn't one here ready to put on so while Ted was working on porting one for me I took the V2 that was on MY car and the new port design we did back to the flow bench. The V2 on my car is different than any other V2 I have seen. It has much more welding on it than I have seen on any others and the runners appear to be cut down lower.

Now back to the flow bench because I wanted to see what the changes we made again to our ported one and this other style V2 were going to flow.

Wilson V2 off my car:
#1, 340cfm
#2, 322cfm (-5%)
#3, 330cfm (-3%)
#4, 325cfm (-4%)
Total CFM, 1317

The BR extra ported intake we have been working on:
#1, 317cfm (-3%)
#2, 313cfm (-4%)
#3, 313cfm (-4%)
#4, 327cfm
Total CFM, 1279

OK, I was happy to see we were able to get ours that balanced with a reasonable amount of work. At this point in the day I was trying to figure out how much more we'd charge for such a great piece of work.

Also note that the Wilson V2 that was on my car had the highest total CFM of any intake I have flow tested. I think this directly cooresponds to the intake still flowing well in the 8,000 rpm+ range. This seems to be the only thing I have really been able to pin directly to the flow bench.

By the time I get the flow testing done the standard ported intake is done and I put it on the car.

The results of the stock ported intake were, 613 whp and 472 ft lbs of torque.

Now to be absolutely honest I expected the stock ported intake to gain in the low and mid ranges since we know we lost 50+ whp on average with the aftermarket intake at 4500. What I didn't expect was for it to be better all the way to 8,000 rpm.

Here is a picture of the three dyno pulls for comparison:

So, a few hundred dollars in flow benching fees, my entire week and most of my sleepless nights this week consumed with flow bench testing, Ted's week spent porting and running around after me, a day on the dyno and what have I learned beyond a doubt? Mitsubishi engineers are NOT stupid.

What else do I THINK I have gathered? Flow balance from runner to runner on the flow bench doesn't seem to effect crap. Shorter runners seem to move the power up OR the difference in the plenum size does it. Kind of knew that already though.

What is the flow bench testing worth? I'd say very little. It does seem that the higher the total CFM is the more likely the engine is going to make power in the high rpm. Is that what it is though? Maybe it's the difference in the runner length doing it and the size of the plenum only.

Wow. I feel like I wasted a week

davidbuschur · Feb 14, 2009, 05:47 PM

The above post is a bit hard to follow as I cut and pasted the testing from our forums so the way it reads is confusing at times. Remember all of that took place over last week and was multiple posts I attempted to put into one and make sense of it all.

Anyone who'd like to read the entire thread in full you can just click here:

http://buschurforums.com/forum/showt...453#post135453

This coming week is going to be more interesting as we are starting a new project.

davidbuschur · Feb 14, 2009, 05:48 PM

OK, I got myself thinking about which way the car was going to be faster. Unfortunately I do not have enough information to determine this. The reason is I don't know how much more the power would have climbed on the V2 and I don't know how much further it was going to fall off on the ported stocker. It pisses me off I didn't run the car out further on the testing but in the past found it wasn't need. By 8,000 rpm one intake has always showed a clear advantage or disavantage over another by then.

I think I can get a good idea though. I went over my log from the 9.04 pass. I used the time on the AEM log to come up with the date I am going to list below. Do you what you like with it but it's something else to consider.

My RS has the two step set at 6700 rpm. I started the timer as soon as the two step turned off. Then I would note the exact time the car hit 8,000 rpm and then again as soon as the RPM spiked from the shift. Then I'd start the time again as soon as the clutch engages after the shift. This is who it looked on the pass.

1st gear the car was under 8,000 rpm for 1.09 seconds
1st gear the car was over 8,000 rpm for .24 seconds

2nd gear under 8,000 rpm for 1.69 seconds
2nd gear over 8,000 rpm for .37 seconds

3rd gear under 8,000 rpm for .74 seconds
3rd gear over 8,000 rpm for .39 seconds

4th gear under 8,000 rpm for 1.93 seconds
4th gear over 8,000 rpm for 2.92 seconds

The car crosses the traps in 4th at 9570 rpm with a trap speed of 159.64 mph.

The total time spent under 8,000 rpm on the pass was 5.45 seconds
The total time spent over 8,000 rpm on the pass was 3.92 seconds.

This gives a total time of 9.37 seconds for whatever reason. The ET on the run was 9.04 I can only go by the time on the log. Kevin may have held it out past the end of the traps, I don't know. It also doesn't take into account the shifts, which I took out completely. Just stating that the time total doesn't match the ET before someone jumps my ass over it.

For me I have to consider that most of a cars ET is set in the first 330' at most. On that end of the track the car is under 8,000 rpm much more than over.

I do wonder being in 4th gear over 8,000 rpm for 2.92 seconds how that power up there effects the MPH.

Just more information.

davidbuschur · Feb 14, 2009, 06:35 PM

Here is the data from the runs of the three intake manifolds being run against each other. I took this from the logs. I am posting 1,000 rpm increments because this is honestly a PITA. Does anyone know how to take 3 datalogs and put the three runs over top of each other, like comparing 3 dyno charts? Is it possible? If so I'd like to take the three intakes and logs and put them over top of each other in a screen shot to post. Otherwise this is going to have to do.

It is very important to note that I did NO tuning of any type during any of this testing. As you will see the runs were so close there was no adjustment needed.

Code:

Wilson V2

              3k      4k       5k       6k       7k        8k 
AIT            59      59       59      59       60        62
02             13.6    13.1    12.1    11.3      11.7     11.6
Ignition       22      23       20      17.8    17.4     17.4
Coolant        176     177      179    181     181      183 
Load          1.6      5.5      19.4   29.3    29.9     30.1  

BR Heavily ported stock
             3k        4k        5k       6k       7k       8k
AIT            57       57        59       59       59       60 
02             13.9     13.1     12.3     11.3      11.6    11.4
Ignition       22       23        19.5     17.1    17.4    17.4
Coolant        174     174       174      174     174     176
Load          1.7      5.7       20.5     29.4    30.1    30.1

BR Standard ported stock
             3k        4k        5k       6k       7k       8k
AIT           57        55        57       57       57      59
02            13.9      12.9     12.3     11.3      11.6   11.4
Ignition      22        23        19.9    17.1    17.4    17.4
Coolant      170        172       172     172     172     174
Load          1.8      5.7        19.9    29.6    30.3    30.6

It would appear to me with the data from each intake being this close that the power differences on the dyno have to be related to runner length, anyone else agree?

RAbishi · Feb 14, 2009, 07:35 PM

It looks like all that hard work paid off. Nice work. Hopefully you can price these reasonably too!

DynoFlash · Feb 14, 2009, 08:58 PM

Fascinating stuff.

The more you progress with the concept of maximizing the low end TQ the faster you are going. A total revolution in tuning.

marrrk · Feb 14, 2009, 09:48 PM

david, a week wasted ? I THINK NOT, its an unfortunate turn of events to bring you to this point. but better now than later, i think... my point is, that week saved me 1500 and i'll bet alot of others as well. i will be sticking with my buschur racing ported/coated intake manifold thank you very, and maybe some additional porting when i come in? thanks for all the relentless testing once again

chris_r_w · Feb 14, 2009, 10:31 PM

David, interesting results.

I have the HKS Kansai IM. How do you feel it stacks up against the mentioned manifolds in real world use, that is on the dyno or the road?

I like the HKS, but would never buy it again, just so much more you can do for the money.

I did feel it lost a significant amount below 4,000rpm, but it's hard to be sure because I installed it along with many other parts that also affect the powerband.

chris_r_w · Feb 14, 2009, 10:34 PM

Quote:

Originally Posted by marrrk

david, a week wasted ? I THINK NOT, its an unfortunate turn of events to bring you to this point. but better now than later, i think... my point is, that week saved me 1500 and i'll bet alot of others as well. i will be sticking with my buschur racing ported/coated intake manifold thank you very, and maybe some additional porting when i come in? thanks for all the relentless testing once again

I'd agree with that 100%...short-term waste of a week, but in the long run it'll effect a lot of potential buyers. If I were to start again I'd definitely be after one of their manifolds.

oldevodude · Feb 15, 2009, 12:28 AM

Short runners=low end power/torque long runners=top end power basic known fact

I guess torque down low to get things going might help the 1/4 mile time. fast out of the hole and pull till the end right??? A lot of normally aspirated cars like my old cobra had butterflies that would open at 3500 to allow a second set of runners to flow more up top. I do not think this is "new news" but great that someone spent the time none the less and all the more

I'm a little drunk so do not mind me

*constant_boost · Feb 15, 2009, 02:18 AM

serious dave, i think i speak for the evo community when we thank you for your hard work and dedication to take the evo world to the next level

dxbevo9 · Feb 15, 2009, 02:45 AM

very interesting stuff!!

scheides · Feb 15, 2009, 06:01 AM

Great post!

davidbuschur · Feb 15, 2009, 08:34 AM

oldevodude, I think you wrote what you wanted to say backwards. The short runners are known for top end power while the long runners are known for low end power.

The week for me was wasted on the knowledge gaining end of things. I saw the flow bench testing that another guy had started for the 1g DSM. I was reall intrigued by it. I thought it was a lot easier than trying to swap intakes back and forth multiple times on the dyno again.

With the recent turn of events for me in the market place on intake manifolds, I also figured it would be a great way for us to figure out how an intake works. The flow bench testing was started and it shocked me how far off the runners were from end to end with the stock one and our standard port job. I then did the V2 and Driven and they were both very balanced from port-to-port. After a lot of work on the flow bench and studying how the ports work we figured out how to balance the intake. Actually by the time all this was done we had our stock intake balanced as good as any, I was pretty proud of that! This HAD to translate to better power..............ends up not.

At the end as you have already read those engineers in Japan at Mitsubishi sure know what they are doing. Our port work to the intakes that we have always offered port match the intake to the head port, change the angle slightly at the head port, the ports are then smoothed out as far back as we can get. Then the throttle body inlet is opened up to 65mm and the entry to the plenum is cleaned up a bit. The intakes are cleaned and such and shipped back. This does increase the flow across the board but it's nothing drastic and the balance from runner to runner is still how Mitsubishi designed it.

While we were doing this the guy who owns/run the flow bench (his name is Buffalo) looked at the intake and he was commenting on the #1 runner being bent and offset and such. I told him I was told that it was like that because the intake had to make room for the brake booster on the right hand drive EVO's. He kind of chuckled and said he bet it was for more than that...............ends up, he is probably right.

Now some other thoughts on this. The plenum on the DI, V2 and the one we did the additional porting on are all larger than stock. The DI I'm going to estimate to be the largest but I'm guessing not my much over the V2 that was on my car. The V2's plenum is greatly increased from all the porting that is done. The intake we ported also increased the plenum size but not by very much. This makes it so you have to consider the intake's plenum effecting the over all curve.

Now the runners are where the major changes are. The stock runners are 10" long on port #1. The runners on the V2 are 6.25 and the DI's are 6.5". The runners in the intake we ported heavily were about 9" long.

It would seem to me that the runner length effects the power band based on the fact that initial testing earlier this year the V2 and DI intakes were very close in power to each other.

It all makes my head spin.

I guess I'll let the cat out of the bag......this week I have a bunch of material coming to the shop to build intake manifolds out of. Everyone else has one so what the hell I may as well attempt it too.

I've got a lot of ideas and a ton of input coming in from a lot of people too who have more educated back grounds than I do.

At this point it appears as though the dyno is the only machine that is going to get the results/power curve where you want it. It also would seem that for 99% of of the stock intake manifold simply cleaned up nicely with a basic port job is the one to use.

I'll have to sell and port a lot of stock intakes at $140 each to make up for this weeks time and flow bench fees! haha

As a side note. Anyone who is getting a stock ported intake from us if you don't want the PCV valve port anymore just let us know and for a small fee we can remove it and plug it for you. Same goes with the throttle body. If you want to eliminate the top vacuum ports or the water feed lines we can remove/weld them off. It just makes it all cleaner if you don't need them. I hate capping things with vacuum caps.

chris_r_w · Feb 15, 2009, 09:39 AM

Quote:

Originally Posted by oldevodude

Short runners=low end power/torque long runners=top end power basic known fact

I guess torque down low to get things going might help the 1/4 mile time. fast out of the hole and pull till the end right??? A lot of normally aspirated cars like my old cobra had butterflies that would open at 3500 to allow a second set of runners to flow more up top. I do not think this is "new news" but great that someone spent the time none the less and all the more

I'm a little drunk so do not mind me

Written wrong way round, but yeah that's the principle; short runners - high end power... longer runners - low end torque.

That's the principle that everyone's always worked by, as you said the butterflies in the runners, or the adjustable length runners on race cars like the mazda 26B (4 rotor) LeMans machine....

...but I have also seen the opposite happen, on my wife's car; a 2L N/A VW Jetta, with a really long IM from the factory I wanted to use a top-mount turbo setup..so I fabbed an aluminium IM using just basic principle's but I did no measuring or harmonic calculations...just a short intake runner for space reasons.

Tried the car before-after on the dyno without the turbo - just the IM and I picked up 15 ftlb of tourque at 2,500rpm's - I couldn't believe it because logic would've suggested the opposite.

Hindsight makes me believe that the original intake me just have been too smooth in it's path towards the head, and that the new manifold promoted more of a "tumbling" effectand therefore better combustion; holding the gas in the air and off the cylinder-head walls resulting in better torque.

Sorry for rambling, it's not even an Evo motor, just trying to offer a different point of view.