Unable to find answer to plenum question despite search...

[CosRush]

Before I post this question, i'd like to say that I have serached through the forums using many different keywords to try and get an answer to my question.

I apologise in advance if the answer is somewhere on here and that I have just missed it.

Information preceeding my question:-

My stock Ford Cosworth Turbo inlet plenum has an incredibly restictive elbow.

images:

Whilst this restriction still does not limit the power able to be produced (upto 500Bhp), and improvements to the restriction in its design can only possibly yield better flow i.e. less friction on the air thereby requiring less Turbo boost pressure for the same volume of air.
I hope that makes sense what im trying tosay.
The incoming air does enter the plenum from the underside just off central so does give a relatively equal flow to all inlet runners.

So.....have things moved on since this original design?

Yes they have. In Sweden where the engines have been extensively used for racing a new design plenum was divised (termed a 'swedish plenum').

images:

The inlet was at one end of the chamber at right angles to the inlets.
The volume was tapered to the rear (i believe in order to maintain the velocity of the air to the furthest inlet ports away from the Throttle, by virtue of the taper increasing the airflow velocity)

The plenum volume itself is also increased. Testing has shown that these flow very well on 500Bhp+ engine but show a slight throttle response drop on lesser engines. I put this down to the increase in volume over the standard.

Now my question........

I have here one of the Swedish Plenums, so if any more information is required in the form of measuremments thats not a problem.

However, this plenum is too big for my particular application. I am looking at having a new one made and would like ot know what factors determine the Taper rate from front to rear???
Taking this top view how would I work the taper required???

I realise that perhaps Maths aside, the best solution would be to make a prototype and test it, make adjustments and retest until a satisfactory result was achieved, but this would be very time consuming and also costly (hence why major manufactures spend a vast amount of money on the designs). Ive seen many 'Homemade' ones and some have shown really very good results far better than the stock item, but I will be straight and say that unless the maths are very simple to calulate the taper, no thought was given to any airflow differences between the inlet runners.
I base this on the very uncomplicated design of those ive seen.

Thanks in advance to anybody who can help

 

[jbthiel]

Remember - indicated boost pressure is the resistance to flow, not total air flow. Like CosRush said - you can reduce the restrictions in your engine, but keep the same cam, turbo, etc. and actually lower your boost while gaining hp.

Also note that the first generation Olds Quad4 intake manifold was an exhaust header in reverse. They replaced it after 2 or 3 years with a traditional cast AL intake - which kept max hp the same but improved torque. I imagine that it was too expensive for mass production.

 

[Deltona]

Yes, I agree with what you are both saying, but I don't think I explained myself fully.

I'm thinking more about the the behaviour of pressurised air, in my head I can see the first wave of charged air behaving totally different to when the whole system is pressurised.
Difficult to explain, but I can see the initial burst (when you come on boost) being like a wave of water
washing along a dry river bed, it favours then easiest route (down the furthest port in the plenum, if you're not careful), but what about afterwards when the whole system is constantly pressurised? I can visualise it being like a full balloon, no matter where you put a hole in it you will get air squirting out. Will these differences in behaviour (if they exist) disrupt any flow bench testing that we may do to show equal port sharing?

Picking up on what you say about using a header in reverse for an intake manifold; Why do we have a plenum on the intake of a high performance car, what purpose does it serve? Why don't we just split the intake into however many cylinders we have and have a pipe to each one straight from the intercooler?

 

[CosRush]

I can see the initial burst (when you come on boost) being like a wave of water
washing along a dry river bed, it favours then easiest route (down the furthest port in the plenum, if you're not careful), but what about afterwards when the whole system is constantly pressurised? I can visualise it being like a full balloon, no matter where you put a hole in it you will get air squirting out. Will these differences in behaviour (if they exist) disrupt any flow bench testing that we may do to show equal port sharing?

I think I see what you are saying but I wonder if its all that relevant. The time between the plenum being unpressuried and then pressurised at say (on my car between 3500-4000rpm) is so quick that any irregularities in the induction stroke at that moment between each runner will surely be over too quick to make any difference to the engine performance since the next induction stroke with the flow more balanced (plenum pressurised) would cancel that out.
As an example on mine at 4000rpm when the engine produces full boost, the inlet valve is opening 33 times a second.........

I feel that in order to test a new plenum design you would also need the inlet section too and measure from the outlet of the inlet (to the head port).
Hope that makes sense..........

Picking up on what you say about using a header in reverse for an intake manifold; Why do we have a plenum on the intake of a high performance car, what purpose does it serve? Why don't we just split the intake into however many cylinders we have and have a pipe to each one straight from the intercooler?

I suspect its down to a combination of cost for manufacture in mass production cars and perhaps little gains v's complicated design/space issues over a standard plenum in modified cars.

 

[Deltona]

" think I see what you are saying but I wonder if its all that relevant. The time between the plenum being unpressuried and then pressurised at say (on my car between 3500-4000rpm) is so quick that any irregularities in the induction stroke at that moment between each runner will surely be over too quick to make any difference to the engine performance since the next induction stroke with the flow more balanced (plenum pressurised) would cancel that out.
As an example on mine at 4000rpm when the engine produces full boost, the inlet valve is opening 33 times a second.........

I feel that in order to test a new plenum design you would also need the inlet section too and measure from the outlet of the inlet (to the head port).
Hope that makes sense.......... "

Sure, it makes sense, but I'm talking about testing the plenum here on a flowbench for equal runner sharing, If you can't replicate real life situations it would warp your results, a flow bench works with only a light vacuum pulling through the ports.
What I would suggest is a small curved piece of metal halfway around the far side of each runner mouth within the end feed plenum to catch the bypassing air and encourage it to go down each port equally by altering its height, angle, or doing away with it etc, you could set these up on a flow bench (If my theory is wrong about boosted air behaviour) before hand, or even full baffles/deflector plates. You could also try out different degrees of taper on the tapered plenums to find out the optimum angle.


Quote:
"Picking up on what you say about using a header in reverse for an intake manifold; Why do we have a plenum on the intake of a high performance car, what purpose does it serve? Why don't we just split the intake into however many cylinders we have and have a pipe to each one straight from the intercooler?"
I suspect its down to a combination of cost for manufacture in mass production cars and perhaps little gains v's complicated design/space issues over a standard plenum in modified cars.

I'm not talking about mass production, but specially designed one-offs like ours, people still follow the same principles using plenum chambers, I have many pictures, but why?
I fancy it will be something about reverse pulses and equal distribution of constantly boosted air, maybe it will be hard to split the one pipe into multiples equally as it will perform differently at different levels of air pressure?

 

[patprimmer]

The plenum is never empty. It is full of air, only the pressure speed and temperature varies.

There are pulsations and constant flow situations going on due to the turbo pumping air fairly smoothly and the inlet valves opening and closing in each port.

Inertia from the air will tend to flow more air to the end cylinder, especially if the plenum curves into the last cylinder.

To stop this, design a straight pipe that is fed from one end and is parallel to the head and goes past all 4 ports then hits a dead end at the other end, say 3" past the last port. Have each runner perpendicular to this pipe and have a bell mouth at the entrance to each runner. That will minimise variations due to inertia effect.

Larger volume of this plenum makes more power, but smaller volume reduces turbo lag. The optimum volume is a compromise between these to conflicts. Only you can decide where that optimum compromise is for your application.



Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[jbthiel]

Once again PatP beat me to the punch

1. The turbo can only spool up so fast - so I don't see the tidal surge happening. But throw this out the window if you have a poorly designed blow-off that goes full open or closed. Then you will get some momentary unpredictable pressure waves - unless you want to spend the next month doing complex CFD analysis!

2. As for the intake header, most of the time higher rpm's are used to get higher hp numbers with non-diesels. This points towards shorter runners which would be difficult to fabricate into a proper header while still providing equal flow and minimal restriction. The Quad4 was gutless below 3500rpm and needed all the help it could get.

3. The vast majority of plenums I have seen are as PatP describes. Constant cross section extending past the last cylinder. I would think if the tapered plenum gave a significant advantage we would see it more. It's probably something the race boys do because of space constraints or to get that last 1.5hp. You might see better transient behaviour because the plenum holds less volum but it is much more design sensitive.

4. Your set-up appears to use a two piece plenum. If you have the time/money/inclination you could design two different covers and see which works best.

 

[CosRush]

people still follow the same principles using plenum chambers, I have many pictures, but why?

Some companies do use them, Isuzu for example.

http://img.photobucket.com/albums/v385/CosRush/34b4_1.jpg

Im very interested to see the results of any design and testing you carry out.
I hope you keep us updated on any progress.

I would suspect the best enviroment for testing though would be 'in use', therefore you would probably be looking at a custom exhaust manifold too since the design of this affects the inlet air too, and if you have a cast manifold you'll struggle to get probes fitted and would need a Wideband lambda sensor on each exh header and also and Egt probe, best optimised if you have mapping software via a laptop. That itself would add upto a lot of money in itself.
Once you were satisfied that you had the airflow right for your own useage you could remove the probes and make use of it.
I cannot see any other way of accurately measuring the airflow in its normal enviroment...........




jbthiel

3. The vast majority of plenums I have seen are as PatP describes. Constant cross section extending past the last cylinder. I would think if the tapered plenum gave a significant advantage we would see it more. It's probably something the race boys do because of space constraints or to get that last 1.5hp. You might see better transient behaviour because the plenum holds less volum but it is much more design sensitive.

I suspect you are right, but most I have seen are infact tapered, and only some of the older designed ones are still parallel. This suggests to me some development improvements both in mass production engines and in motorsport.
Probably as a result of the latter (in the same way that F1 benefits from Aerospace design).

 

[jbthiel]

CosRush, the Isuzu manifold has some of the same design issues that we have been discussing. The middle two runners are a straight shot for the incoming air and are shorter than the outers. The outer two also have more aggressive bends. The plenum is not very big, so it won't help high RPM power. My friend deals with manifolds similar to this when he hot rods tractor engines (updraft 4 & 6 cyl moters) - he says the distribution sucks but he has to use them to meet the rules.

In the end this is like almost every other engineeing question - DEPENDS!! There are so many competing variables that you will never have a perfect design. We all have to live with good enough.

 

[CosRush]

jbthiel
In the end this is like almost every other engineeing question - DEPENDS!! There are so many competing variables that you will never have a perfect design. We all have to live with good enough.

Thats exactly why I think this is such a complicated area to do in a low budget DIY way. Manufacturers spend millions of pounds designing similar components, which is why I think anybody who considers making one themselves will either get a rough compromise (which many would live with) or get lucky and have a good balanced flow or worst case scenario baldy flowing one with very uneven air distribution.

Ive already shown this to Deltona, but it does make a good representation of the difficulties.
This is the final Ferrari F430 inlet plenum test results, and it clearly shows that despite Ferrari's design,testing and very large budget they did not achieve totally equal balance to all runners.

http://img.photobucket.com/albums/v385/CosRush/F430_press-intake-plenum.jpg

The only way IMO is testing testing testing......

 

[Ataloss]

Do you know what the range of pressures was in that Ferrari test (that is, what's the delta P between the 'red' and the 'blue' zones)? Without numbers on the colour scale it's tough to determine just how uneven the distribution is. If the red is 100kpa and the blue is 90kpa, it's much less severe than if the red is 100kpa and the blue is 10kpa.

I'm just curious since a lot of analysis programs automatically scale the colours to span the range of numbers, whether it's 1 unit or 1,000,000 units.

Thanks for sharing that pic.
Bob

 

[brngrhd]

i would also like to know if the pressure difference is caused by the taper in the manifold or if that is helping it? it looks, to me, like taper is causing more pressure difference. would internal velocity stacks make a difference in the flow in that manifold, making it more even? if you had a manifold like pap was talking about and had with the plenum parallel to the head and runners parallel to that would you have the same problem with the end cylinder getting more air? would stacks help that?

 

[patprimmer]

Stacks won't help for several reasons, not the least of which is they won't change the pressure profile inside the plenum. Also required tuned length varies with air pressure as the speed of sound changes.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[brngrhd]

so is it more because of the taper in the plenum? i have been reading a lot on the forums because I'm making a manifold for my car too. i don't want any issues with one cylinder getting fed more then another. the manifold will be equal length runners so all i really need it plenum size and shape. and from what i have read it sounds like 50-60% of total engine volume.

 

[Deltona]

CosRush,

If you are going to use the Swedish style plenum then consider the angle at which the throttle body presents itself to the plenum, you can angle it so that it blows towards the first few runners rather than straight in, see here:

http://s103.photobucket.com/albums/m127/Deltona_II/?action=view&current=IM000609.jpg

This engine is a 2 litre and is pushing out some serious power (in excess of 700bhp)

 

[CosRush]

Deltona

CosRush,

If you are going to use the Swedish style plenum then consider the angle at which the throttle body presents itself to the plenum, you can angle it so that it blows towards the first few runners rather than straight in, see here:

http://s103.photobucket.com/albums/m127/Deltona_II/?action=view&current=IM000609.jpg

This engine is a 2 litre and is pushing out some serious power (in excess of 700bhp)

Thanks for that

Ive decided though for the time being to use the standard plenum until such a time that I have available time and money and tooling to design a better plenum for my engine.
I personally would not be happy until I knew the airflow was as good as (but preferably better than) the original, and unfortunately I have run out of time to get involved now, so on the 'back burner' for the time being.

I hope you make some progress with yours however, keep us posted of any progress, especially with respect to any results you find

 

[Lou Scannon]

patprimmer, the last time I checked, the speed of sound was a function of temperature only. But I'm sure you knew that, and meant to type 'temperature' instead of 'pressure' in your post!

 

[patprimmer]

Temperature and pressure are related, but you are correct, it was an oversight.

The changing pressure, and therefore the changing temperature in a turbo application makes pulse tuning to difficult to bother with.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[womble]

Thats exactly why I think this is such a complicated area to do in a low budget DIY way. Manufacturers spend millions of pounds designing similar components, which is why I think anybody who considers making one themselves will either get a rough compromise (which many would live with) or get lucky and have a good balanced flow or worst case scenario baldy flowing one with very uneven air distribution.

Granted you can go to the nth degree in design detail... But for DIY I don't think it's that hard. If you go for a front entrance plenum design that has the volume you require, tapering down to a minimum height above #4 runner (1.5xrunner_id minimum)you should get very good runner to runner distribution - bare in mind OEMs may allow up to 5% variation between runners. OEMs do use steady state flow bench testing for validation also, so if you have a tuner near you that has a flow bench you can ask if they can flow test it for you. Guy Croft Racing Engines has a flow bench if you get stuck.

 

[patprimmer]

My recommended method would be to make a plenum in 2 pieces so it can be split for internal access. Take your best shot with the volume on the small side.

You can then easily experiment with volume by using spacers at the split line.

You can read the plugs, measure exhaust gas temperature or even run oxygen sensors on each exhaust pipe to determine cylinder to cylinder variations. As the plenum splits, you an access the bell-mouth of each runner to rework it. Then rework the bell-mouth on each runner inlet to get the maximum charge you can get with uniform distribution. Resist the urge to lower the good cylinders to the lowest common denominator and try to improve the rest to equal the best.

While flow benches can be a useful tool, they are constant flow and therefore do not tell the full story, like variations in adjacent cylinders due to charge robbing due to firing order. They also do not give any help re plenum volume so long as the plenum CSA is enough to supply steady state flow to the furthest cylinder. Flow benches do not help predict the effect of pulses in the manifold and do not help in predicting the effect of volume on the magnitude of these pulses or oscillations.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[jbthiel]

CosRush - I don't think that you can do much worse than the production pretzel of on intake you have! Pat's recommendation mirrors mine above. You already have a split plenum - just make a new cover for it and give it a try. You can always unbolt it for fine tuning. I also like Pat's idea of adding a spacer to try different plenum volumes.

If you want to get the most bang for the least money - modify your existing cover by milling off the throttle body mount and welding the hole shut. Then mill a long slot and weld on a tapered inlet section. You can reuse the throttle mount cutting off the flange and welding it onto this section.

I also like the dual inlet idea you showed - if you have the room for it. This takes a lot of the equal flow problems out of the equasion.

http://i103.photobucket.com/albums/m127/Deltona_II/DeltaThrottlebodies3100.jpg

------------------------------------
An off subject item - how do add the quote box in your reply? If looked and can't find how to do this.

Thanks. IceStationZebra