The Science of Dump Pipes

[cpitts]

Actually, we are OFF TOPIC.<snip>
What I asked about was what is the science behind a cast collector, vs welded steel

Ooops. Interesting debate this one. One specifically targetted at heat retension in the first 30cm of the exhaust, post turbo, to allow faster spool an preserving air speed and flow.

The thermal properties of a ceramic coated welded setup are far superior to that of a cast lump, but that cast lump costs just a fraction to make.

Had this discussion with a couple of people previously. The general theory goes that the cast outlet will build up heat and remain warmer/hotter than a standard dump pipe, helping air-speed out of the turbo and reduce lag on gear change/off throttle moments.

NOW, the down side is that there's a MUCH greater time required to heat up heat up the massive lump of steel that is the cast collector.

SO to combat this you look at ceramic coating, heat treating (HPC coating), wrapping (fibre, silicon, copper, foil). The arguments seem to rage around the benefits of HPC coating the inside v's outside of the pipe and the affect it has on the heating of the pipe metal itself to preserve heat in the pipe metal to, again, help with lag reduction and airflow out of the turbo.

So much more to discuss on this, coyote, would be happy to hear your views on this, I have a few and they seem to align with yours on external coatings of the pipe, but not necessarily on top of a cast collector.

Ironically, I seem to be pointed ultimately at getting a cast collector simply due to the fact it has the splitter that I'm after, not for the pre-heating aspect. I suspect I'll need to heat wrap and/or ceramic/HPC coat the outside of it as well to keep that critical heat load in there and avoid excess heat in the engine bay from the big lump of steel.

Both the AVO and Phat GT pipes for the 2.5 have a similar (though different ) collector design, including the splitter.

Last time I looked, Phat didn't offer a pipe with a splitter. If he does, by all means point me to it, I may be very interested.

[GTBR46]

Ooops. Interesting debate this one. One specifically targetted at heat retension in the first 30cm of the exhaust, post turbo, to allow faster spool an preserving air speed and flow.

Actually, this is precisely the point I was trying to tease out. I have read about this in general on a few forums when I was considering boosting my Mx5 a few years ago. I could recall the discussion, but not the specifics. I would have thought that the period defined by "MUCH greater time required to heat up heat up the massive lump of steel that is the cast collector", would be a relative thing. It might take "5 times as long" for example, which is 5 times a not very long period. The point is, once up to operating temperature, it is supposed to aide performance, and unless you are just doing a quick squirt to the shop and back, it means for the majority of the time, its meant to be better than steel. I have no fact to back this up, just read about it previously, and thus posed the question. Im already ducking, but it is not meant to be a criticism of any existing design, just a discussion on materials. Of course, a shit design with a better material is not as good as a better design with...lower performing materials.

[senator]

You can get almost anything cast really cheaply in Chinese foundries.

The thermal properties of a ceramic coated welded setup are far superior to that of a cast lump, but that cast lump costs just a fraction to make.

+1

Don’t forget the issues of precocity and metal composition of the casting

IMO the only reason why avo use a casted mouth is to be able to apply the same mouth design to a number of different dump pipe applications….. eg wrx, gt etc etc casting is just a cheap way to make this part at the end of the day.

I seriously doubt the ability of any casting house to maintain the tight tolerances required to achieve casting the splitter in the mouth too.

NOW, the down side is that there's a MUCH greater time required to heat up heat up the massive lump of steel that is the cast collector.

.

and the last thing you would want is a large mass of hot metal close to transmission

A there any dump pipes out there that you can buy off the shelf with a spitter fitted to suit our libs???

btw not much more for a splitter since the wastegate opens in a mirrored position when compared to previous wrx turbos

[Johnb]

I spent 12 months ummingg, erring and ahhing over dump pipes. Here's my two cents worth based on car enthusiasm, past engineering studies and an attempt to put it in plain english.

As far as the science goes you need to do some serious reading on gas dynamics and gas flow theory. If your keen here is a sight to read http://eom.springer.de/G/g043350.htm

Attempt at plain english explanation
There are two basic designs for after market dump pipes:

1. the open or bell mouth like that from MSR Engineering and AVO or
2. the twin tube dump pipe like that from CES racing or the Phat design,

In Australia we are restricted to a 3 inch dump in both design as our cars are RHD and the steering knuckle catches on a 4 inch design. In the USA 4 inch open bell mouths are popular. Both designs can have variations with or without different splitter configurations. Everyone has an arguement for the product they are selling.

With exhausts the complication is that the gas is hot and released under pressure from the turbo. A "good exhaust design" should allow the gases to escape the turbo quickly without melting the rest of the car or burning objects as you go past them (no sarcasm intended).

So basically what gas dynamics and gas flow theory tell us is that to improve gas flow rate you need to allow the gas to expand quickly, flow in straight lines, eliminate objects that cause turbulence and allow the hot gases to cool at a rate which does not slow the gas velocity to the point that it creates a back pressure. Back pressure can also be created by reducing the diameter of the pipe or introducing barriers to gas flow such as poorly designed mufflers and cats or acute bends in the pipe.

Both the open bell mouth and and twin tube designs allow the gases to expand quickly and flow at a rate that allows the gases to cool without causing back pressure (much quicker than the Subaru design). As a result lag time is reduced as there is less resistence to the compressor wheel turning due to the decreased back pressure and increased rate of escape of the gases from the turbo. This also has benefits in increasing power and torque. And introduces new issues around turbo spiking, more rapid spooling and higher operationing temperatures which can lead to failures of various components for a range of reasons. This is why its necessary to get a retune and look at fitting a better (bigger) intercooler whether it be TM or FM.

Ceramic coating (inside and outside) helps with the dynamic properties such as allowing the gases to cool at the required rate and reducing the transfer of high levels of heat to the engine bay, gearbox and car floor - heat transfer can make the intercooler work harder, shorten engine life (usually bearing failure), shorten gearbox and clutch life, boil power steering fluid (has a lower boiling point than brake fluid) amongst other things.

The advantage of the Phat design is that it has a pipe for the exhaust and a pipe for the waste and these two gas streams can travel for longer than the bell mouth design before converging. The exhaust is the higher volume gas so it has the bigger pipe.

In addition the barrier between the two pipes is narrower in the Phat design than other designs so there is less creation of eddies in the gas (turbulence) and so less introduction of shear forces (there will always be some). There are other designs similar to Phats but the bit I like is that it does not have a splitter extending back toward the turbo. The issue here is that in some designs the waste gate can catch on the splitter and restrict the opening of the waste gate forcing the gases sideways thus introducing turbulance and shear forces. There are reports on wrx forums and legacygt of waste gates breaking in these circumstances.

Now the other three reasons I chose the Phat dump are:

1.it weights less than dumps with cast iron collectors,
2. Its very well made, and
3. It looks good (that AVO dump pipe is ugly).

cheers

[Arith]

Great write-up John! That really explains it in layman's terms and puts it in perspective. For me it confirms my confidence in Phat's dump pipe.

[coyote]

Just look at any car that has had a cast iron dump for a period of time. The area surrounding the dump gets damaged by heat radiating from the dump. Then look at one with a coated steel pipe, there is no such damage. If you want, measure it with a temp probe (I did many years ago). Now ask if the heat didn't get radiated into the surrounding area, where did it go? The answer is of course that the heat stayed in the gas, which is exactly what we want. Cast dumps absorb heat out of the gas, then radiate it into the surrounding parts. Any statements you read to the contrary are just marketing bullshit dressed up as a technical claim.

Splitter vs No Splitter vs Bellmouth vs Divorced Pipes ... is a completely different discussion.

[coyote]

Yep, well said JohnB.

Had I read your post first, I would not have bothered penning mine.

(and I'm glad you gave credit to CES for the twin pipe design, rather than the hordes who went and ripped it off)

[cpitts]

Senator, not sure what you mean by procosity, did you mean porocity?

Physical metallurgy » Casting » Metallurgy » Porosity

A major problem in castings, porosity is principally caused by the shrinkage that accompanies solidification. Molds are designed to feed metal to the casting in order to keep it full as solidification proceeds, but, if this feeding is incomplete, the shrinkage will show up as internal pores or cracks. If these cracks are large, the casting will be useless. If they are small, they will have relatively little effect on the properties.

Another cause of porosity is the presence of gas-forming impurities in the liquid metal that exceed the solubility of the gas in the solid. In such cases, solidification is accompanied by the formation of bubbles as the gas is rejected. To eliminate this problem, gas-forming elements must be removed from the liquid before casting. Bubbling an inert gas such as argon through the liquid before casting is one means of doing this; vacuum degassing is another.

This is just a quality control issue. I know AVO have gone through a heap of different foundrys trying to get other items cast to their desired tollerences.

I also agree that a cast or other bolt on head makes mfg of dump pipes SO much easier as you just have 2 or 3 different heads and bolt it up to a standard cat/dump design. So much more flexible for client and provider in event of upgrade.

Yes, heat transmission from the cast head into the engine bay and surrounding items is also a major issue/concern that I see (other than time to heat up to operating temps). Hence if you head down this path, I'd (personally) suggest some sort of exterior coating or wrapping for this part to keep the heat IN rather than radiating out from the cast head after it's been at peak load/heat. A heat shield will also help keep things under control as well, hence why there's one for the factory turbo/dump.

To try and address Hamish71's question:
Hot air occupies a bigger volume than cold air.
Air is the hottest at the back of the turbo.
A lot of the cooling occurs in the first 30cm of the exhaust
SO, you're kinda looking to get a bit more diameter up top and taper down from there.

Also, air moves faster when it's warm, so once the heavier cast steel warms up, in theory it'll stay warmer, keeping air speeds up in that critical first 20cm after the exhaust port when you're not on peak power. Giving you better throttle response when using the car under on/off peak situations. It's probably arguable though that in urban stop/start travel whether the cast outlet would actually be able to reach a good temp so it could actually achieve this goal. (hence the need for wrapping, etc to keep heat in) The principal is exactly the same as heat wrapping headers, extractors, up pipes and other exhaust components, and this has been proven time and time again.

Ultimately, all designs have trade-offs. Some are designed for race, others for sports street, others for basic replacement/upgrade. There's no 'one fits all' version. However, certain aspects will benefit all situations like the splitter, cat design and placement, mouth shape and pipe diameter.

Is there off the shelf versions of these pipes with splitters? Well, AVO has them but I think they actually weld the splitter in as they adjust length to suit the turbo it's bolting up to, so the term 'off the shelf' is more a case of 'upon request'.

Think that covers some of the discussion and no doubt starts a lot more.

[cpitts]

<snip>2. the twin tube dump pipe like that from CES racing or the Phat design, <snip>

John and Coyote,
Clearly you've both been down this discussion before, and I'd be interested to hear your view on excess pressure being forced back into the turbo or not being allowed to escape from the waste gate, leading to the mentioned boost spiking issues that's specific to this split pipe dump setup.

Clearly if you're not letting the waste escape, boost's goign to spike which results in more power, so ruling out the obvious side effects of higher boost levels, more power (as a result), extra heat and heat related failures, what info have you found about the above phenominon that exists with this twin pipe setup.

It's been a common setup for a long time with consistently presented issues. I don't think it's due to turbo overspooling or over running, I think it's a failure to let components do their job properly. Is there evidence to demonstrate otherwise? Prodrive would disagree with you, and who am I to argue with them!

[senator]

Senator, not sure what you mean by procosity, did you mean porocity?

This is just a quality control issue. I know AVO have gone through a heap of different foundrys trying to get other items cast to their desired tollerences.

Thanks for the spell check........

Btw save me the text book explanation and no it’s not a quality control issue, material choice comes to mind.

Trust me casting is not something you can earn from text book ………….. I live and breath this crap, and actually know the offshore supplier avo is currently using, after they failed with a few US and OZ suppliers.

The casting world is not that big

[cpitts]

Senator, wasn't meant as a jab (the porosity thing) just trying to make sure we're on the same page.

I've also got no doubt that the casting scene isn't that large as volume importing would make it more efficient to have a fixed number of importers rather than a raft of mum and dad sized SMB's. Not wanting to educate you on a scene I know very little about, just passing on direct statements made to me by AVO about other casting examples they've had.

From re-reading your post comments re tolerences required to cast in the splitter; I believe they weld it in afterwards so it's a local tollerence thing rather than a casting requirement. (As stated in my earlier post).

[smythie]

Air is the hottest at the back of the turbo.
A lot of the cooling occurs in the first 30cm of the exhaust

The intention of the exhaust turbine is the convert as much of the heat energy and pressure in the gas into kinetic energy (rotation of the turbine wheel). Thus the exhaust gas will be cooled, gas pressure will reduce and volume will increase as it flows through the turbo. If none of those occurred, the turbine wouldn't have any power to drive the compressor and you wouldn't have boost.

In short, exhaust gases are hottest in the combustion chamber, and drop off all the way through their flow to the atmosphere. The biggest temperature gradient is in the turbine housing itself.

[cpitts]

I have no argument with you there Smythie. Cylinders are where the most heat is, everything after that has cooling devices/systems attached to them, including the turbo!

I was just referring to the stuff that bolts onto the back of the turbo.

[Boxer]

Clearly if you're not letting the waste escape, boost's goign to spike which results in more power, so ruling out the obvious side effects of higher boost levels, more power (as a result), extra heat and heat related failures, what info have you found about the above phenominon that exists with this twin pipe setup.

It's been a common setup for a long time with consistently presented issues. I don't think it's due to turbo overspooling or over running, I think it's a failure to let components do their job properly. Is there evidence to demonstrate otherwise? Prodrive would disagree with you, and who am I to argue with them!

Hmmm commonsense would dictate that if boost spikes were due to the wastegate not being able to do its job properly with a twin pipe setup - that the phenomenon would be exhibited moreso with the OEM flange plate Subaru serve up.
This would pose far more restriction than any turbulent cause from a twin pipe dump. More likely the boost spikes if apparent are due to faster boost onset and lack of control over the wastegate from the ECU. Again this is evidenced by a tune or EBC fitment being able to fix it. I have not seen Prodrives musings on the subject, so can't comment on their twin pipe opinion. But having used CES three times would put a lot of faith in his engineering and performance exhaust knowledge. Still the only guy I have ever seen gain HP from an NA TypeR without changing headers and he did it be reducing pipe size, while everyone else made them bigger.
I do note much of your comments are referenced to 'what I have heard' or 'what I have read' or ' what AVO told me''

[cpitts]

I do note much of your comments are referenced to 'what I have heard' or 'what I have read' or ' what AVO told me''

It's not my job to design exhausts, nor do I run an exhaust shop.
I know what I've used, I know what others have used from their first hand experience. I've talked and discussed many options and reasonings behind the many aspects of certain parts of the turbo Subaru setup that interest me.

You gain knowledge by reading, talking and doing. Some vendors are prepared to discuss things with you (eg: Vin from STI Docklands (and BGT prior to that), Steve Coates and the boys @ AVO, Adam at TuSpeed, even Phat on here!) others are not. I don't have the time to put against buying, evaluating and testing each pipe on the market, that'd just be a gross waste of time and money. Hence the need to research.

Don't you research stuff and talk to other owners before you go and buy something to modify your car/house/bike/etc?