Manifold Design and Turbine Housing - expensive lesson learnt

[Steve]

I have a HKS 3037 on a custom manifold. The manifold is tuned length highmount. Originally the turbine housing as 0.61 AR, which brought 1 bar of boost on at around 3000rpm, and hit very hard - too hard, but torque was flat to redline and gave a reasonable 256rwkw at 1 bar with a very 'quick fix' tune.

At this stage, I was going to get cams fitted, with a 1.6mm head gasket to drop the CR and allow a safety margin for higher boost. I decided at the same time to get the 0.87AR turbine housing fitted, as I was concerned that the small 0.61 housing was going to choke the top end.

Initially on the road the car was very crisp off boost, showing much better characteristics than before the cams and head gasket went in. All was good. I decided to give it a 'gentle' thrash.

Lag, lag, and more lag - the car was a dog. I couldnt think what could be the problem, unless something mechanically was not right. The new setup didnt make good boost (plant you in the seat) until nearly 6000rpm! Nothing at all happened until over 4500rpm. I almost cried.

I sent the car off to be tuned. Fortuneately for me I managed to secure the services of an excellent tuner. He tuned the car for 2 days - on the road. Idle is excellent, power is very smooth and very good off boost - alot better than before. The lag however has only been moved 800rpm lower (thanks to -6deg on the exhaust cam, and very good tuning).

the problem as Richard sees it (he is doing the tuning, and also was responsible for JMSs Red R33 - which still has the original motor in it, just spun a bearing with 2 bar boost on a cold engine, long story) is that I have a manifold with long runners, that collects into one outlet. This is very different to the way the Japanese have made their manifolds for street use turbos. They (japanese) use very short runners, which arent tuned length, and according to Richard are designed to get very high exhaust gas velocity before it hits the turbo, which a tuned length 6 into 1 will not do as well. The tuned length manifold is very well made, just doesnt match the turbine and application I have. Richard is convinced the problem is the manfold isnt creating enough exhaust gas velocity to use the 0.87 AR housing, which isnt too big for the engine, but is too big for a tuned length manifold and this turbo. He has seen the japanese manifolds, and they work very well at getting gasses into the turbine housing very quickly, use some crazy rocket science which comes from extensive R&D - therefore allowing the use of bigger turbine housing.

Now I have a bit of a quandry - do I buy a Jap made (designed for this turbo) manifold, redo the dump, wastegate, afm and intercooler to turbo pipework, or go back to a small housing? I am not sure the smaller housing will be ideal for top end, but it does wonders for bringing in boost early and spooling the turbo quickly. But I dont know which option would be optimal for response and top end. I want it all if I can have it:D I am a bit concerned about back pressure, as I am running inlet an exhaust cams - too much back pressure will rob power very quickly, especially as I have a bit more overlap now than stock cams. (oh and the VVT still helps alot with off boost response, even with the cams fitted - much nicer to drive with it connected) I dont want to go changing things again only to have to redo them again, its not a cheap exercise (r&d is expensive:()

Either way, I have learned a very valuable lesson, that I thought I would give anyone who is at the point where they are considering turbo and manifold upgrades something to consider, maybe in a bit more depth than I did (and I did spend alot of time reading and researching, just didnt ask the right people I suppose:))

The runner size and design of the manifold obviously has a massive effect on what housing to use, and also how boost is being made - dont underestimate the proper design of a manifold. Something to think about before deciding which manifold and housing to use. And, dont underestimate the amount of science that goes into the design of a Jap manifold - they make them that way for a reason.

Rule of thumb, longer runners, smaller turbine housing, shorter runners, bigger housing.

May have a cermamic coated tuned length RB20/25 manifold on the market in the new year:D Either that or a 0.87 HKS GT30 turbine housing:p

[meggala]

a great read steve thanks for sharing the japs also tend to have biiger exhausthousing and laggier cars cause they spend alot of time in heavy traffic and off boost. if you want that low down power I would be tempted to go for the the smaller exhaust housing and put up with the lack of top end if it exists. I also think the compression drop was part fo the cause for the lagg to be exadurated (spelling) as the cars stock compressin is very high and I think it was required for safety. put hte smaller exhaust housing in and see what happens you might get a surprise and it might be quite good b4 you go and change exhust manifolds.

meggala

[Steve]

Glad you enjoyed it, I will be definately doing the housing first, just a shame that I already sold my 0.61 (hindsight) then take a look at top end and decide whether or not to change the manifold again.

the off boost is very, very good - much better than with stock cams and no head gasket, so although I think you are right, it does effect off boost and spool, it would appear that the cams (and adj cam gear) more than compensate.

Can never tell exactly when you do more than one mod at once, but as the head was coming off, I thought it would save a few dollars.

[BOOSTD]

I can understand how this causes a problem. The gas will be travelling slowly up to the .8 A/R scroll then accelerated at an enormous rate through the scroll. The resulting pressure difference between the manifold and turbine housing would cause turbulance increasing back pressure. I guess continuous smooth acceleration is what ur after to reach max flow and velocity. So in this theory no wonder matched manifold/turbo kits kick serious butt. Wonder what the formula looks like to work this out..........scary

Great read Steve.

[Steve]

yeah, probably good for top end (which it appears to be) - and is what traditional tuned length exhaust manifolds do.

The jap ones for the street tend to go 3/2/1 or blend the pipes into eachother on the way, and have a pipe before the turbine flange - Richard suspects this acts to accellerate the gasses before they get to the turbine housing.

I always looked at these and though 'it has to be better with minimal interference before the turbine', as it would reduce back pressure there fore promote scavenging. Most of the research I have done pointed in this direction too unfortuneately.

At least I know now, tuned length great for drag or use a small housing if you want it on the street. I cant wait now to get the smaller housing on and see how much the top end suffers.

[BOOSTD]

exactly how the adjustable cam gear works. Can't have the best of both worlds. I might try -6degrees if my turb is laggy. Did that setting loose you much up top ?

[KamikazeR33]

very good read as my exhaust manifold is being made this w/e, so i will inform the guy who is makin it to make sure he does it this way

[Steve]

I dont know what the top end will be like if I move the cam forward, might have to ask that question of Richard, as I didnt drive it before and after he started adjusting, and it hasnt been on a dyno yet.

[Bl4cK32]

Great read Steve, thanks for the heads up on the manifolds. Sh*t though about the situation though...

could you sell off the manifold, dump, and buy a setup that suits the .8 ?....fark its still gunna cost either way.

Guess you need to decide how much more yer gunna spend

[Steve]

Yeah, dollars:)

I was thinking it might be a good back up for if I ever decide to do a 3L. Much torqueyer (if there is such a word) engine, and bigger capacity would make a big difference. I am also wondering if it would be possible to mod the manifold, shorten the runners and put a collector collector before the gasses hit the turbine housing. Also, it 'may' be reasonable with an inbetween size housing on there. A few different options, so I think I will tread a little more carefully for now.

It does get expensive when you factor in dump, wastegate pipe, and all the pipework off the compressor on top of the cost of the manifold. I am not too sure I would like to sell to anyone unless they were doing a 3L or after a drag setup.

I will just try the next size down in housing for the time being, I may be lucky and strike a nice balance between runner length/turbine housing/backpressure - fingers crossed.

[Warpspeed]

Very nice read Steve.

One thing I discovered was that too much valve overlap also does what you are experiencing. It feels like it really wants to go off boost, but as soon as you floor it, it dies in the bum.

You might be right, or you might be looking at the wrong end.

If its not too hard, can you put the stock inlet cam back in ? Leave the long runners and the big exhaust housing fitted, and see if the boost threshold drops back down.

If it does, it might be an exhaust reversion problem, and not caused by low exhaust gas velocity. What inlet cam did you finally decide on ?

[Steve]

I am running with a 256/8.8. I was thinking it might be reversion too, but with the VVT disconnected it ran worse, wouldnt it run better with it disconnected if overlap was causing the problem?

Also, it doesnt die inthe bum, it just takes ages for the boost to come in. I still have the stock inlet cam, might be worth a try, esp if changing the turbine housing doesnt do much.

Pity I still havent got my head around this Dyno 2003.

[Freebaggin]

Interesting stuff. Steve, i thought you were going for 264°/9.15mm HKS exhaust cam with the stock inlet? What made you change to the 256°/8.8mm?

So just confirming the changes you made from 256rwkW @ 1 Bar (with broad spread of torque) to what you have now:

*exhaust cam change

*greddy plenum

*0.87 exh housing

*1.6mm gasket

Good luck with next stage. You might be right though, good 3L setup.

[Roy]

Sorry to hear things didnt go to plan!

Id just go back to the smaller exhaust housing. Bump up boost a little and be happy with say approx 265-275rwkws, i know i would be more then happy wiht those numbers, especially with boost hitting so low in the rpm and so hard. Can i ask what power the bigger exhaust housing is making? Perhaps plot some rpm, points and equivelant power?

A little left field, but what have you ended up using as far as inlet plenum is concerned, Greddy, custom, std?

The design of my trust manifold is as you say 6-2-1. Wouldnt have thought that the equal length would be the problem, more the runner diameter and blend points. What OD and wall thickness is your manifold runners?

[Steve]

I was always planning on 256 in and 264 exh. I have seen this combo used with very good success. It is interesting that the smallest inlet cam HKS make would have such a detrimental effect. You got it right, inlet and exhaust (256/8.8 x 264/9.0) adj exh cam gear, 1.6mm head gasket, 0.87 turbine housing and a trust plenum.

Also, I am pretty sure Merli is using 256/264, and he has good boost from just under 4krpm, and he is using a HKS 3040 with a larger turbine housing from memory.

It just doesnt add up. I am pretty sure the cams are good, esp as disconnecting the VCT didnt result in an increase of power, as it should being that the cam would be further advanced to 4500rpm resulting in more overlap - please correct me if I am wrong.

Seriously, I have been going over and over this, as has Richard, there is only one unknown, and it is built more like a jap drag manifold than a jap street one - everything else is a good thing, it has been proven.

I now have a trust plenum, I didnt want to mess with another aust made job. Too much R&D in trust to ignore, as it appears there is in HKS manifolds.

Roy, what make is your manifold? Wanna sell it:) or even better trade:D got a really nice tuned length job, ceramic coated and works killer with smaller turbine housing;)

Once I get a smaller housing on and see how it goes, it will become more apparent, if it still is sluggish, look somewhere else, but at this stage, best method of attack is to start looking at any unknown quantity first, I hope:)

[Clint32]

Steve do you have any pics of the manifold on the car, wouldnt mind a look. I have a blitz manifold you could borrow and try if you like as i wont be needing it for a while,

[Merli]

Bah... I just wrote out a long explanatory post, but you seem to know it all already, so I don't want to waste your time.

Okay, here's what I think. If you do a 6-1 collection at the end of the manifold, you have 6 different exhaust gasses coming from 6 different directions, all fighting to go into that one tube at the end.. Result? Turbulence, and hence backpressure and (even greater) reduced gas velocity. Coupled with your extra long runners, you've slowed the exhaust gas quite considerably.

Whereas if you take a look at the HKS exhaust manifold, from memory, it's a 6-2-1 design, or something like that. If you merge 3 pipes into 1, you can have a much tighter grouping, with exhaust gasses going in the same general direction... With the 6-1, you have gasses from cyl 6 and cyl 1 hitting each other from nearly opposite directions!!

The 6-2 grouping is done very close to end 2-1 grouping, so you have to look twice to notice it, but it's there.

[GTR-Ben]

The HKS setup seems to be very good.

I have a Subzero manifold for my RB30 setup.. it is rather odd looking and I can only hope it dosn't cause any problems like stated here.

Here are some pictures comparing the manifolds.. the photos were taken on a phone so sorry if it's hard to see anything at all..

HKS Manifold. If you zoom in or something you can see that it goes from 3 into 2 just before it goes into 1.

Subzero Manifold. Certainly not tuned length and goes 6-1. I might have to get one like the HKS manifold made..

[Steve]

Sorry Clint, I dont have any photos. I am still going to get my oil cooler fitted pretty soon, so you will be able to check it out when I come looking for oil lines;)

thanks for the offer of the manifold, I will keep it in mind when I have determined for sure it is the manifold that is the cause of my woes (when I change the turbine housing again)

Merli, what manifold are you using? which cams and turbine housing? And lastly, where exactly do you get good (say around 1 bar) boost being made?

Cheers

Steve

[Steve]

GTR Ben, those runners look a very similar lenght to mine (sub zero), but a bit hard to tell, also mine is mainly pipebends, and goes down and back up if you know what I mean.

I like the large wastegate pipes on the HKS manifold too, I think perhaps the wastegate is somewhat overlooked sometimes, it can do more than control boost, the right size can also do alot to relieve back pressure from what I understand.