Calling Max_Torque - manifolds and turbos question

Calling Max_Torque - manifolds and turbos question

Author
Discussion

turbotoaster

645 posts

171 months

Saturday 20th August 2016
quotequote all
sl0wlane said:
Still working out how to do intercooling and exhaust (no, I'm not doing a charge-cooler! I'm going air-air)

Umm that's about all the relivent bits I think.
have you seen my exige air to air setup?

sl0wlane

Original Poster:

669 posts

192 months

Saturday 20th August 2016
quotequote all
I have Lee, and I'm eagerly waiting to see your results... Hopefully you can do some proper data logging of inlet and outlet temps on the intercooler so we can see the delta.

I might not be able to do the same as, unlike your car, I will have a hot manifold and turbo right in front of the intercooler if I use that position, being that the K20 has the exhaust ports on the either side of the head to the Rover as you know.

I'm waiting for Concept Racing to open (appears they close and go on holidays for August)... Then I'll talk to them about manifold and intercooler design and placement to see if they can help (http://www.conceptracing.co.uk)



sl0wlane

Original Poster:

669 posts

192 months

Saturday 20th August 2016
quotequote all
Ive said:
in general a plenum should have a certain minimum volume that is usually a multiple of each cylinders displacement.
I remeber it being like 10x to avoid power losses from pressure drop in the chamber as the intake stroke happens.
so for a 2.4l your want some 6 liters inthe ideal world. in reality you get away with much less before significan losses occur.
My old KTM Duke 2nd edition came with a tiny airbox from the factory. 609cc single attached to a like 2l airbox is no good. Increasing plenum volume to infinite by fitting a open air filter. no other changes than a bigger mainjet released plenty of top end power.
The length is entirely dependent of the rpm band. I have sned you the spread sheet with some empirical fomulas in it.
you need cross section and lengh from mouth to intake valve. Then check where the 2nd, 3rd and 4th harmonics are.
if you strive for 7000 rpm peak torque, you want your effective lengh tuned to the 2nd or 3rd harmonic at that rpm. 1st would even be better, but I doubt you can and want to fit such long snorkel to your engine.
My old plastic plenum on my supercharged K would really pick up again past 7800 rpm. This fitted to the 4th harmonic of this geometry. In between 5000 rpm of the 3rd harmonic peak and the 7800 of the 4th there is the "valley of death". here the refelcting sound waves actually reduce the pressure at the port during the intake stroke. You could feel it and see it from the fuel values to maintain AFR and from the ignition timing I could add before knock. VE was dropping.
The shorter VVC plenum shifts things upward, so I am less affected by the drop past the 3rd harmonic peak. Still, past 7000, power does not rise anymore as VE drops.
If I would fit ITBs with shorter runners than the VVC plenum, VE would drastically increase at high revs. This can be seen from dyno data. otherwise identical spec NA engines,m even the same cam timing, generate some 160HP on a plenum at 7000 and some 180HP on ITBs.

have you calculated the peak rpms for your intake set-up? what length is the intake path from mouth to intake valve? What is the (average if conical) cross section?

The only reason the Honda k20 engines get away with such a short plenum regarding the low and mid rpm torque is their reduced cam timing and lift to help VE in those load and rpm ranges. At the top they need the short wide runners to be able fill the cylinders effecitvely.

if you check some online articles comapring various plenums on K20 engines with dyno charts, you can nicely see the influence of the length of torque and power. as this influence is large pressure independent, you can orient yourself on thos lenghts regardng your intended torque curve.

Edited by Ive on Friday 19th August 12:18
Thanks Ive, most helpful! The good thing about the Jenvey manifold is that I can change the length very easily by changing the spacers (currently 50mm long), from your info above it looks like shorter might be better (I had not taken into account the VTEC at lower RPM).

I need to measure the manifold but also the length of the ports to the valves in the head to get a handle on this fully.

Another question... What RPM should I target?! So far I have been working on 6000rpm, as this is right in the middle of the power band, but I guess on a pure race car you would target perhaps higher ~7000rpm (I guess then it comes down to the gearbox ratios as well).... Such an interesting and multifaceted subject! smile

Gunt

11 posts

89 months

Monday 10th October 2016
quotequote all
Hi , New here but i said i'd just chime in

i see what you are trying to achieve but you are watching the wrong details way with too much emphasis , those twin plane plenums are always only ever used on engines where restrictor's are required WRC etc and or to get around a serious original design issue , ie where there is a retro fit to a section of original two piece manifold .
watching all the guys who achieve serious out puts they are hovering around 2.5 time the engines cc for manifold volume .
if you have ever seen a proper n/a turbo conversion , you can take the n/a value and add 1 bar .The new charged out put is exactly double as its atmospheric value has now just doubled [ on a well designed set up ]
you are asking the question and trying to guess where you want peek tq , this is a honda I V-tec motor there should be no peek , say bar a few where spool comes on until spikes are full controlled , to me you should be looking at the near perfect n/a graph dropping compression , optimizing a turbo reasonably long 4 branch twin waste gate would help [ not needed on a honda ] you should see a 4-5k power band peek tq to bhp even at 700 so 5-600 bhp will be no problem , my biggest question is what fuel are you aiming to use as this will dictate more than the rest , putting this power through 2 wheels , you wont want peek tq just linear progression always predictable

sl0wlane

Original Poster:

669 posts

192 months

Tuesday 11th October 2016
quotequote all
Gunt said:
Hi , New here but i said i'd just chime in
Thanks for posting smile

Gunt said:
i see what you are trying to achieve but you are watching the wrong details way with too much emphasis , those twin plane plenums are always only ever used on engines where restrictor's are required WRC etc and or to get around a serious original design issue , ie where there is a retro fit to a section of original two piece manifold .
Hmm, I agree a good portion of them have inlet restrictors... But some don't, including I believe the current BTCC cars (which are boost limited but do not run a restrictor as far as I can see)... Another example is Max_Torques car, which I don think has a restrictor either - hence asking him to comment.

Gunt said:
watching all the guys who achieve serious out puts they are hovering around 2.5 time the engines cc for manifold volume .
if you have ever seen a proper n/a turbo conversion , you can take the n/a value and add 1 bar .The new charged out put is exactly double as its atmospheric value has now just doubled [ on a well designed set up ]
I agree, but I'll put it to you that this is the result of having a good EMAP vs IMAP ratio, e.g getting on for 1:1, rather than most turbo cars, including factory cars that are more like 2:1 or worse (certainly worse if running higher boost than original design intended).

Gunt said:
you are asking the question and trying to guess where you want peek tq , this is a honda I V-tec motor there should be no peek , say bar a few where spool comes on until spikes are full controlled , to me you should be looking at the near perfect n/a graph dropping compression , optimizing a turbo reasonably long 4 branch twin waste gate would help [ not needed on a honda ] you should see a 4-5k power band peek tq to bhp even at 700 so 5-600 bhp will be no problem
We have ended up going for around 24-27" equal length, wastegate priority manifold, constrained by the packaging in the car... I have hopes of the final figures starting with a 5 or if really lucky a 6 smile

Gunt said:
my biggest question is what fuel are you aiming to use as this will dictate more than the rest , putting this power through 2 wheels , you wont want peek tq just linear progression always predictable
Fuel is going to be pump "super", this will be a road going car, I don't want to be limited to fancy fuel. As such I can't see it would be wise to go any further than around 22psi, maybe a touch further but will need to see what the knock threshold looks like once we are into mapping.

Also totally agree, I want a nice progressive power curve, so I have an electronic throttle, boost by gear and throttle input and traction control to help with that... I also have a sequential box, so keeping it in the power band should be easier (if a little scary).

Wheel spin will no doubt be a problem...

Dummy engine going in the car and fabrication of the manifold, intercooler and various plumbing starts on the 17th biggrin

Gunt

11 posts

89 months

Tuesday 11th October 2016
quotequote all
depending on all i usually work things out here [ compression power band ] and get to run 24 psi on 95 oct , leave your compression a bit high and youll make super power less lag wider power band , stepped manifold helps too

227bhp

10,203 posts

127 months

Tuesday 11th October 2016
quotequote all
You'd be surprised how little high CR makes, it's mainly used for economy and emissions in the real World.

sl0wlane

Original Poster:

669 posts

192 months

Wednesday 12th October 2016
quotequote all
Gone for 9:5:1.. So not mega low, but a good drop from the factory ratio

Gunt

11 posts

89 months

Wednesday 12th October 2016
quotequote all
from anything we have done and from what ive seen , high comp will give more in most cases than people tink and combined with fuel oct limit it an often give more as no more boost can be applied with out an issue due to fuel , it allows you to run less boost for more power less chance of det less heat to deal with , we never build anything under 9:1 on 95 , and currently trying 10:3 on a honda with 99 id go 9.8:-10:1 , proper combustion chamber shape will help any issues

sl0wlane

Original Poster:

669 posts

192 months

Saturday 26th November 2016
quotequote all
Well we have got this far so far... waiting on a box of Inconel....


sl0wlane

Original Poster:

669 posts

192 months

Wednesday 30th November 2016
quotequote all
Just a bump in the slim hope Mr.Max_Torque sees this and takes pity on me smile

I'm just down to exhaust manifold runner length, the inlet runners are easily adjustable later, but fabricated Inconel exhaust headers not quite so...

Any help would be really really appreciated! I honestly google the hell out of us every night (and have done for months and months) and still don't have a satisfactory formula for determining the exhaust manifold dimensions... inlet theory and N/A exhaust theory quite easy to find and apply... perhaps I just use the same calculations?

Pretty please with a cherry on top smile

Edited by sl0wlane on Wednesday 30th November 07:04

stevieturbo

17,229 posts

246 months

Wednesday 30th November 2016
quotequote all
sl0wlane said:
Just a bump in the slim hope Mr.Max_Torque sees this and takes pity on me smile

I'm just down to exhaust manifold runner length, the inlet runners are easily adjustable later, but fabricated Inconel exhaust headers not quite so...

Any help would be really really appreciated! I honestly google the hell out of us every night (and have done for months and months) and still don't have a satisfactory formula for determining the exhaust manifold dimensions... inlet theory and N/A exhaust theory quite easy to find and apply... perhaps I just use the same calculations?

Pretty please with a cherry on top smile

Edited by sl0wlane on Wednesday 30th November 07:04
Given crappy cast manifold designs spool fast and make good power.....really, dont worry about it too much. Even worse log manifolds also work

So what sort of extreme are you working to ? How much space do you have to create a fancy design ?

And really...there are bound to be thousands of similar size engines making similar power goals to use as a model. There will be no re-inventing the wheel here, anything you want done you can be sure has already been done many times both cheaply and very expensive. How much difference between the two in terms of performance ? Probably not a huge amount in a lot of cases.

the 500hp power goal is quite modest. So keep tube sizes on the smaller size, if you can go twin scroll and if you can equal length all good too.. But really....if you could not make 500hp these days, something is wrong.

"Optimising" lengths etc....would surely involve head flow details, camshaft details and other factors and probably lots of either simulations or dyno testing. If that isnt within your budget for your particular setup, dont get too worried about it.



227bhp

10,203 posts

127 months

Wednesday 30th November 2016
quotequote all
I think we're going round in circles here, all the salient points have already been mentioned.
One thing I haven't spotted though, what fuel are you going to use?

sl0wlane

Original Poster:

669 posts

192 months

Wednesday 30th November 2016
quotequote all
stevieturbo said:
Given crappy cast manifold designs spool fast and make good power.....really, dont worry about it too much. Even worse log manifolds also work

So what sort of extreme are you working to ? How much space do you have to create a fancy design ?

And really...there are bound to be thousands of similar size engines making similar power goals to use as a model. There will be no re-inventing the wheel here, anything you want done you can be sure has already been done many times both cheaply and very expensive. How much difference between the two in terms of performance ? Probably not a huge amount in a lot of cases.

the 500hp power goal is quite modest. So keep tube sizes on the smaller size, if you can go twin scroll and if you can equal length all good too.. But really....if you could not make 500hp these days, something is wrong.

"Optimising" lengths etc....would surely involve head flow details, camshaft details and other factors and probably lots of either simulations or dyno testing. If that isnt within your budget for your particular setup, dont get too worried about it.
Thanks for the comments but you are missing the point somewhat old bean.

I have an opportunity to firstly learn, secondly build a properly optimised system, I'm not trying to throw this together.

The fact that I'm having it fabricated in Inconel could also maybe indicate that I'm not looking for the cheapest / easiest option.

If you look at the above photo you can see I have plenty of space for pretty much any design size / length required.

I'm asking because I want to understand the science and be able to try my hand at designing the best possible solution.

The calculations do indeed require a number of inputs you mentioned and more, but if designed correctly and as close as I can to ideal it avoids many iterations of design / build / dyno... so surely doing the theory, if I can and have the opportunity makes sense?

sl0wlane

Original Poster:

669 posts

192 months

Wednesday 30th November 2016
quotequote all
227bhp said:
I think we're going round in circles here, all the salient points have already been mentioned.
One thing I haven't spotted though, what fuel are you going to use?
The points have been covered, but not by Max... frown

Fuel... initially just "pump" super / 99, 1000cc Injectors should hopefully be big enough for initial power goal... but I'm eyeing up a secondary fuel rail for the plenum for methanol injection later on perhaps.

stevieturbo

17,229 posts

246 months

Wednesday 30th November 2016
quotequote all
sl0wlane said:
Thanks for the comments but you are missing the point somewhat old bean.

I have an opportunity to firstly learn, secondly build a properly optimised system, I'm not trying to throw this together.

The fact that I'm having it fabricated in Inconel could also maybe indicate that I'm not looking for the cheapest / easiest option.

If you look at the above photo you can see I have plenty of space for pretty much any design size / length required.

I'm asking because I want to understand the science and be able to try my hand at designing the best possible solution.

The calculations do indeed require a number of inputs you mentioned and more, but if designed correctly and as close as I can to ideal it avoids many iterations of design / build / dyno... so surely doing the theory, if I can and have the opportunity makes sense?
Then you need to find somewhere where you can give them all the relevant data, have them run many computer simulations for various designs and take it from there. I'm sure this would only cost 5 figures. Or start building test designs and put them onto your engine and see what works best for whatever setup you're running at the time. I'm sure places like Prodrive etc could oblige

or a more practical approach as I've said....take a look at what the countless thousands of other builds have done that already do what you want. Look at what the fastest cars are using....and you'll find it's what I've already said.
Dozens of different designs that all work....or maybe hundreds....or more.

k20erham

372 posts

125 months

Thursday 1st December 2016
quotequote all
Compromise is the word closely followed by intended use, I've looked after/ rebuilt k20/k24 creations from both NBR and Mountune running all kinds of management on regulation fuels and top fuels, it's a fantastic platform to start from but each path to an optimum in one direction takes you away from another, the optimum for all in unison is the compromise I mention along with the cost, there are not many transmissions that will take what you seek if driven hard, and if driven hard don't forget to factor in 100hr overhaul intervals.
Please keep us posted, and as already advised try not to reinvent the wheel.

227bhp

10,203 posts

127 months

Thursday 1st December 2016
quotequote all
sl0wlane said:
227bhp said:
I think we're going round in circles here, all the salient points have already been mentioned.
One thing I haven't spotted though, what fuel are you going to use?
The points have been covered, but not by Max... frown

Fuel... initially just "pump" super / 99, 1000cc Injectors should hopefully be big enough for initial power goal... but I'm eyeing up a secondary fuel rail for the plenum for methanol injection later on perhaps.
I think you'll be det' limited by octane rating and CR, limited to what (max bhp) only time will tell, but i'd guess at 550, max 600 FWBHP, WI is the option I would take if or when you are

You will find he either doesn't know, or doesn't have a day to spare typing out a thesis on the subject as the resulting equations would be huge, with many variables (compounded even further by Vtec) that even a supercomputer would struggle with and you also haven't listed a precise enough spec to be able to do anything with. There isn't a clear B&W answer as you can gain at low rpm and lose at high rpm, vice versa and anywhere in between or outside of those parameters.

You won't get very far by simply using Google, I would post up or search on some of more dedicated forums such as Eng-tips, Speedtalk and other turbo forums. Larry Widmer would have a take on it, but difficult to get him to talk unless you crossed his palm with silver.
Ultimately the answer is to get a Dyno Sim package and do it yourself or pay someone else to run it, I know some members on here have them, but again lack of precise spec would be a problem.

Knowledge or answers are gained by the application of time and money.

stevieturbo

17,229 posts

246 months

Thursday 1st December 2016
quotequote all
227bhp said:
I think you'll be det' limited by octane rating and CR, limited to what (max bhp) only time will tell, but i'd guess at 550, max 600 FWBHP, WI is the option I would take if or when you are

You will find he either doesn't know, or doesn't have a day to spare typing out a thesis on the subject as the resulting equations would be huge, with many variables (compounded even further by Vtec) that even a supercomputer would struggle with and you also haven't listed a precise enough spec to be able to do anything with. There isn't a clear B&W answer as you can gain at low rpm and lose at high rpm, vice versa and anywhere in between or outside of those parameters.

You won't get very far by simply using Google, I would post up or search on some of more dedicated forums such as Eng-tips, Speedtalk and other turbo forums. Larry Widmer would have a take on it, but difficult to get him to talk unless you crossed his palm with silver.
Ultimately the answer is to get a Dyno Sim package and do it yourself or pay someone else to run it, I know some members on here have them, but again lack of precise spec would be a problem.

Knowledge or answers are gained by the application of time and money.
Exactly, even if it isnt the answer he wants to hear.

Whether knock will be a limitation....I dont think he has specified a CR ? There is a current fascination with running high CR's ( ie close to 10 or sometimes 10.x )...all good and well if you have the fuel to cope with it.

For mere mortals in the real world on pump fuel etc....high 8's is still perfectly fine and will easily and safely allow his goals to be met without the need for silly mega expensive fuels. Obviously if the budget does extend to expensive fuels, than that is something you can exploit.

227bhp

10,203 posts

127 months

Thursday 1st December 2016
quotequote all
stevieturbo said:
Exactly, even if it isnt the answer he wants to hear.

Whether knock will be a limitation....I dont think he has specified a CR ? There is a current fascination with running high CR's ( ie close to 10 or sometimes 10.x )...all good and well if you have the fuel to cope with it.

For mere mortals in the real world on pump fuel etc....high 8's is still perfectly fine and will easily and safely allow his goals to be met without the need for silly mega expensive fuels. Obviously if the budget does extend to expensive fuels, than that is something you can exploit.
Listed as 9.5:1 - rocket fuel territory. Some OEMs are running 10:1, but with low boost, very clever cooling systems, DI and well developed knock control it is possible, not all things which you can apply to older generation engines.
You can of course make it work using other methods, it just needs a lot of knowledge, money and the resulting added complication.
Anyhow, we're off topic....