Exhaust Design

[landroversforever]

While I'm waiting on some bits for the project I'm looking at the exhaust design. So does anyone on here have any detailed knowledge of exhaust design?

The Mercedes OM606 I'm using has sort of oval ports which are 615mm2 each.

Now as I'm building a compound turbo setup I want to get the most from the exhaust that I can. Do my manifold tubes need to match the area of the port? and therefore be 48mm  diameter or do they need to be smaller? 

At the moment I'm struggling to find any decent exhaust design tech stuff! 

 

Thanks in advance.

[dangerous doug]

of you want to get the most out of it your runner should match the port. I have matched the port size to the runner size on all of the heads I have ported and it's always been an improvement. If you're building your own tubular manifold you Might want to take into account scavenging and presurewave tuning but that can turn into a headache and you might be forced into just making what fits. I have a link somewhere for a manifold design kit, it's pricy but you basically clip plastic sections together to make your manifold and then transfer that into tube sections.....I'll dig it out 

[dangerous doug]

http://www.icengineworks.com/

 

[landroversforever]

Ah I saw that a while ago.... crazy expensive  Wonder if I can find a design to 3D print one!

17 hours ago, dangerous doug said:

of you want to get the most out of it your runner should match the port. I have matched the port size to the runner size on all of the heads I have ported and it's always been an improvement. If you're building your own tubular manifold you Might want to take into account scavenging and presurewave tuning but that can turn into a headache and you might be forced into just making what fits. I have a link somewhere for a manifold design kit, it's pricy but you basically clip plastic sections together to make your manifold and then transfer that into tube sections.....I'll dig it out 

I'm not sure how much the pressurewave and scavenging will affect stuff as its a turboed engine? Do they still apply? The exhaust is going to be at a much higher pressure than in a n/a engine as its going to be driving two turbos.

 

[dangerous doug]

I have a very limited knowledge on it bud. It's a complex subject effected by many things that I do not pretend to understand but It will still have an effect as you can potentially be sending a positive wave up a runner to the valve and effectively chocking that valve. I would have thought the pressure is all  relative so Doing this will aide breathing and allow it to rev higher (dont interpret that as moving power up the rev range).

People have got around this by making manifolds with big runners to lose the wave in but that big runner needs more pressure to spin the turbo and you're back to square one.

Do you need a new manifold for compound turbos? I thought the outlet and inlets were linked on the turbos And Keeping a stainless manifold from warping might be another challenge if your running compound setup

[landroversforever]

I don't know how choked the valves are going to be anyway with the high pressure driving the turbos.

Well I don't have to have a new manifold as I can make an adaptor to the turbo flange, but I see it as an opportunity to make a better flowing design. The standard manifold looks like this on the OM606:

The turbos are linked as follows. Exhaust goes out of the port into the HP (small) turbo, the exhaust side of that (and the wastegate) flow the exhaust to the LP (large) turbo. On the intake side, the air goes into the LP turbo, through into the HP turbo and into the engine.

I've been talking this morning with a couple of other friends and Ian from IRB Developments about the intake runners. the ports on the 606 are a 28x46mm rectangle with a 14mm rad added to each end. The surface area of that is equivalent to a 2" pipe . Our consensus at the moment is to gently take that down to a diameter similar or a touch larger than the standard manifold but just make it with a better flow than the standard layout. 

One of the guys also came out with the rule of thumb of 1" dia per Litre for the exhaust so a 3" system from the (last) turbo back does seem about right from my inital research.

[Bowie69]

What do dieselmeken etc do for manifolds?

The merc one doesn't look that bad in all honesty, and in a tight engine bay looks excellent for packaging.

Is that 1" per litre for NA engines or turbo's as well? Personally it sounds like a crock, as pipe volume goes up with the square of the radius, so not linear. A 5l rover V8 does not need a 5" exhaust. If he told you anything else, disregard it

Exhaust sizing will depend on air flow through it, which will depend on engine size, rev limit and whether it has turbo's on it or not, and what pressure they run at.

[landroversforever]

Depends on the power levels and the turbo being used. DPUK make a manifold adaptor to a big single like an HX35 and you can buy off the shelf manifolds for the bigger turbos. I don't think will work for me as I don't want a big laggy single.

I'm wanting to improve on the standard manifold to get the best from the smaller turbo. I want it to start boosting from about ~1500rpm I recon. and it doens't look like the standard manifold will do that at the moment. My aim is to get the drivability of a V8  Not sure how close I'll get but I'll give it a shot.

That was for turbo diesels. I don't know where the rule came from but he's a clever chap normally so I'll let him off if others think its rubbish

So some basics on the engine:

• 3L
• Revs to I think 5.5K
• Twin turbos
• max boost needs to be about ~3Bar ~44psi? 

[David Sparkes]

Quote

I want it to start boosting from about ~1500rpm I recon.

I achieved something like that, but it was all done by controlling the flow of air to the waste gate pneumatic actuator.

But that was on a BMW straight 6 cylinder. I don't know your OM engine at all, so the same approach may not work.
On the BMW engine the  waste gate does NOT vent excessive pressure to the atmosphere; the 'waste gate' is actually a bypass valve, allowing exhaust gas to bypass the turbo and go straight down the exhaust pipe.

The basic characteristic to be understood, and accepted, was that the waste gate is never fully closed while a gear is engaged and the engine is delivering enough power to move the vehicle, even if it's only at tickover. The standard pneumatic actuator ensured the waste gate started to crack open at very low pressure, thus bleeding off some of the exhaust flow from the turbine. The modification was to stop that premature bleed.
Basically, pressure valves were fitted in the pressure feed to the actuator. One was held totally closed, so the waste gate never moved and was as sealed as it could physically be, given its imperfect seating arrangement. At a pressure determined by experimentation, and with temporary pressure gauges connected to both the inlet manifold (turbo outlet pipe) and the feed to the actuator, the closed valve 'snapped' fully open, when the waste gate opened to bleed off the exhaust gas flow from the turbine.
The second pneumatic valve, in series with the first valve, was initially fully open, it started to close under the influence of the pressure seen at the turbo outlet pipe. This valve set the actual maximum pressure seen in the inlet manifold.
Thus this arrangement gave me a steeper rise in manifold pressure, AND the option of an increased maximum pressure.

Once set up the control system was stable, and required no further attention.
It wasn't my original idea, the guy at the Australian online magazine did the development, but on a petrol engine. Given the petrol emphasis in that country I don't believe he appreciated the potential it liberated in passenger car diesel engines, of the sort used in Continental Europe and the UK.

As I say, I don't know the twin turbo OM engine, but I'm tempted to think that IF the waste gate is pneumatically controlled then adding finer control into the pneumatic control circuit will be a hell of a lot cheaper than creating or modifying manifolds.

Regards.

[FridgeFreezer]

1 hour ago, landroversforever said:

I want it to start boosting from about ~1500rpm I recon. and it doens't look like the standard manifold will do that at the moment. My aim is to get the drivability of a V8

Surely with no/low boost until 1500rpm you're not getting the V8 drivability? I'm with Mr Sparkes on this, exhaust is not super critical compared to other factors. Also, Mercedes have more time & money than you and they designed the original manifold like that...

[landroversforever]

36 minutes ago, David Sparkes said:

I achieved something like that, but it was all done by controlling the flow of air to the waste gate pneumatic actuator.

But that was on a BMW straight 6 cylinder. I don't know your OM engine at all, so the same approach may not work.
On the BMW engine the  waste gate does NOT vent excessive pressure to the atmosphere; the 'waste gate' is actually a bypass valve, allowing exhaust gas to bypass the turbo and go straight down the exhaust pipe.

The basic characteristic to be understood, and accepted, was that the waste gate is never fully closed while a gear is engaged and the engine is delivering enough power to move the vehicle, even if it's only at tickover. The standard pneumatic actuator ensured the waste gate started to crack open at very low pressure, thus bleeding off some of the exhaust flow from the turbine. The modification was to stop that premature bleed.
Basically, pressure valves were fitted in the pressure feed to the actuator. One was held totally closed, so the waste gate never moved and was as sealed as it could physically be, given its imperfect seating arrangement. At a pressure determined by experimentation, and with temporary pressure gauges connected to both the inlet manifold (turbo outlet pipe) and the feed to the actuator, the closed valve 'snapped' fully open, when the waste gate opened to bleed off the exhaust gas flow from the turbine.
The second pneumatic valve, in series with the first valve, was initially fully open, it started to close under the influence of the pressure seen at the turbo outlet pipe. This valve set the actual maximum pressure seen in the inlet manifold.
Thus this arrangement gave me a steeper rise in manifold pressure, AND the option of an increased maximum pressure.

Once set up the control system was stable, and required no further attention.
It wasn't my original idea, the guy at the Australian online magazine did the development, but on a petrol engine. Given the petrol emphasis in that country I don't believe he appreciated the potential it liberated in passenger car diesel engines, of the sort used in Continental Europe and the UK.

As I say, I don't know the twin turbo OM engine, but I'm tempted to think that IF the waste gate is pneumatically controlled then adding finer control into the pneumatic control circuit will be a hell of a lot cheaper than creating or modifying manifolds.

Regards.

The OM normally uses a vacuum controlled wastegate on the turbo. Most factory waste gates are internal rather than external ones venting to atmosphere. Can't stand the silly noises the external ones make. 

How was it all controlled in the end then? Something like a Rasbery Pi or Arduino? Boost control is something I've not really got round to looking at yet, but there are various things I want to control to get the best out of the compounds. Stuff like a bypass for the LP turbo until the HP has started producing ~0.5bar maybe.

[landroversforever]

38 minutes ago, FridgeFreezer said:

Surely with no/low boost until 1500rpm you're not getting the V8 drivability? I'm with Mr Sparkes on this, exhaust is not super critical compared to other factors. Also, Mercedes have more time & money than you and they designed the original manifold like that...

The specs for the OM are 174bhp @ 4400rpm and 243lb/ft at 1600rpm. That's really not that dissimilar to the 4.6 from my research? 

The other option I'm half considering is a blower instead of the first  turbo.

[geoffbeaumont]

I think what FF meant was that you would still have very little torque below 1500RPM. The big drivability advantage of a V8 over most turbo diesels is that you have plenty of torque right from idle.

[landroversforever]

I'm not sure on the figures as most of the drift lot are only aiming at peak power, but the E class it came out of picked up really nicely from idle. The modified injection pump should make it pick up even sooner too. Will have to see how it performs! 

[David Sparkes]

1 hour ago, landroversforever said:

... How was it all controlled in the end then? Something like a Rasbery Pi or Arduino?

No, no, nothing electronic at all, just the Norgren pneumatic valves. One was rebuilt with a different spring (softer, IIRC) to get the active point mid range.

However, if the OM6 uses vacuum control then my experience doesn't exactly crossover.

Good Luck.

[landroversforever]

I should have added that I'm not using the factory type vacuum control, I'm sticking with the pressure system.

Have you got a link to the valves? 

[landroversforever]

If I can do what I want mechanically then that would be great :D. I was otherwise wondering about using an Arduino.

 

[Bowie69]

OK, I'll stick my neck out... I think you are probably over-thinking all this.

It sounds like you are quite wedded to a compound setup, but have you thought about just running a single VNT? 

A compound setup gets awfully complex and loads of huge pipework in a small engine bay, yet you are going quite conservative on the power front?

Another option to consider may be a twin scroll turbo? They have better pickup at low revs.

Modern turbo's are much better that what was available back in the 90s so I am sure you can improve things from stock very easily.

Just thinking out loud

 

[dangerous doug]

^^^^^^^^^

what he said

 

too much BHP  and a 3 link front is asking for trouble anyway

[landroversforever]

All good comments!

There are a few reason why I want to go the compound route. Some reasonable, some 'tart-ish'. Firstly I want the power and availability of power that compounds will give, the big Turbo I want for my power level won't have the spool characteristics I want. And visa versa for the small turbo. Annoyingly the popularity of the Mercedes engines has balooned since I decided to go that way so I basically want to go one better! Which I know is probably a bit childish but oh well . My goal is a lovely flat torque curve as early as I can get it and 400bhp. Which I don't think id get even with a decent modern turbo. And another from the tartish perspective is so I can say I've got twin turbos in the pub etc. I'm also enjoying the challenge of the design/research and having the final product that I've built.

It shouldn't be too bad with the 3-Link, its never going to stick to the road like my Mini does but with the X-Deflex and the OME shocks it should handle reasonably well. The wheel and tyre combo at 57Kg a corner is going to swallow some power too when I need to spin and clear them.

I'm mainly looking at the Holset turbos as that's what the Merc lot have experience with. They also seem to pretty efficient from what I've read. Twin scroll ones also only seem to be the larger ones from what I've seen.

Any thoughts on a VNT turbo that can spool low and flow the air needed for the top end Bowie? 

 

 

[Bowie69]

Holset and Garret both do VNT/VGT versions of their turbo's.

I'm no expert on sizing, but a GT37 should be good for 400BHP, I think you spent some time staring at compressor maps recently?

[David Sparkes]

4 hours ago, landroversforever said:

I should have added that ... I'm sticking with the pressure system.

Have you got a link to the valves? 

I've got better than a link, I've got links to the later re-runs of the original articles.

But first a point of explanation.
I was surprised to find the Autospeed.com site still existed, because no new articles have been written for some time. For example, Julian Edgar's last Blog post was August 2016.
Because I was a subscriber, I got an email announcing every issue, so I know that Issue 731 was the last issue, but I now cannot find his valedictory article contained in there.

You will see that JE was, and probably still is, something of a perfectionist, and this sometimes leads him to go OTT.
A case in point is the air pipe connections in the circuit.
'Suck this egg' .... Air pressure in 'pounds per square inch' is EXACTLY what it says; 'Force per unit area'.
Say the pressure is 30 psi, if this pressure is supplied to a membrane that has an area of one square inch, then the membrane has to be mechanically strong enough to support a force of 30 pounds.
Reduce the membrane in size to an area of one quarter of a square inch. Maintain the pressure at 30 PSI. What mechanical force does the membrane have to withstand?, Yes, one quarter of 30 pounds.

If you use quality small bore rubber piping then large threaded connections are NOT required.
I found Volvo cars used such piping in their screenwash and headlight wash systems, and as most of them were large cars, there was always a lot of piping to be recovered from each example in the scrapyard.
Don't forget the tee junctions, as they are also robust enough to withstand the small forces involved. If you need a straight through connector then a small self-tapping screw, with a smear of Blue Hylomar on the threads, will be perfectly adequate to seal off the redundant leg.
Nor are clamps required; the natural grip of the rubber pipe is more than adequate. The only place I used clamps was the one area where I couldn't access the joints at the roadside, this was the connection from the turbocharger outlet to the waste gate actuator.
Over several years I never had the rubber pipe perish, crack, or harden.
Remember that these pipes do not 'flow' air, they simply transmit pressure, so a small internal volume can only help speed changes from one end to the other.

The other 'master tip' I will pass on concerns gauges. Forget automotive 'turbo' gauges; go instead for 0 to 30 gauges sold to monitor the line pressure in a gas welding set (there may be other examples that give 270 degrees rotation for 0-30). Forget digital gauges; you are dealing in transient pressures and trying to observe the road as well, most of the time 'passing 10 heading to 15' is probably the best you will get.

Once you have a system set up you will not need both gauges, just the one showing manifold pressure, so you can go for something prettier then, if you insist. The gauges I used were so sensitive that the manifold pressure gauge would 'rattle', all the time. I reasoned that it was pressure waves created as the inlet valves opened and closed, causing the mechanical parts in the gauge to chatter against one another. My cure was NOT to JE standards. I took the pipe off the gauge, inserted a square section matchstick in the round section pipe, reckoning that the restricted area would dampen the force of the pressure waves, while still letting the average pressure be visible through the open sections left when you put a square peg in a round hole. It worked.

I used the two Norgren valves, I didn't feel the need to go for the Italian jobbie. :-)

Part One, and Part Two.

Regards.

[FridgeFreezer]

I realise you know what you're doing but you can't put too much store in what the drift boys are doing (or other racers), their use case is the opposite of what you want in a LR.

[30psi]

Hi guys am following this with interest and plan on doing a similar swap. Have been on the Superturbodiesel forum talking to the merc boys.

I know a Swedish company/bloke on Facebook making a compound turbo om606 manifold, and am similarly torn between the number of options, like landroversforever I want around 400hp for a number of reasons, however in reality I want the driveability and low-down torque as well as the high-end boost.

I've come up with a few options that I'd love feedback on (if anyone's that bothered haha)

Engine - OM606 with Dieselmeken 603 8mm pump and tubular manifold from KKD or Dieselpumpuk

- Big single turbo i.e. Holset HX40 Super

- Compound setup

- Big VGT (no idea how to control it and can't rebuild)

 

Gearbox - Arguably more interesting - Must be able to take the power and couple with LT230

- Adaptor to R380 (not reliable with ~400lb ft)

- Merc 6 speed manual (available from many clk models) - Requires custom flywheel, clutch plate, pressure plate etc. No kit off the shelf as far as I know

- Nissan 5 speed manual conversion from 4x4fabrication.co.uk - proper job but not sure how expensive

- Merc 722.6  5 speed automatic, incredibly tough and meant to be very reliable, can add electronic controller (paddle shift in a 110!!) - Not sure about autos. I do know Rakeway sell an adaptor to fit the rear of this 'box to the LT230

 

Sorry for the thread hijack and anyone interested probably has this information already.

 

 

 

 

[Daan]

I still think 2 small VNT turbos, both from 3 cylinders each , then merge them together to go through the intercooler would be very much the way to go. making two identical 3 way manifolds is much easier than 1 off 6 way manifold. The turbos can be small and cheap, and will spool up really quick. Compound means you end up with a large and really large turbo in the engine bay. It would be very expensive, very difficult to package, complicated piping route and probably do little at low RPM.

 

All you need to do then is sort out the VNT control.

 

Daan