Twin scroll.

[witek]

Hi guys,
Been lurking a while and finally decided to slowly embark on my first gas turbine. I'm going to take the simplified route for my first engine and attach a combustion chamber to a turbocharger using compressed propane as fuel.

I've sourced a turbo and have a general question regarding the turbine housing. I've read in at least a couple places that having a bifurcated or 'twin scroll' turbo will complicate things with a gas turbine. I've not heard any reasons as to why exactly it would complicate things, rather just read to avoid them. Does anyone know why this would be suggested? The turbo I've sourced is a AiResearch T04 (same as garrett) and it's got twin scrolling in the turbine housing.

Cheers

[finiteparts]

Hi "witek",

There are several reasons why the twin scroll housing are not ideal for these engines.

The first one is the durability issues. The gas flow coming out of the combustor impinges directly on the dividing portion, which would be like you sticking something directly down in the flow path. The leading edge of the divider will have the flow stagnating on a portion that will be subjected to the full total temperature of the gas flow and the heat conduction path is very limited at the top and bottom of the inlet. This means that that center divider section will be much hotter and trying to thermally expand because of this, but the outer portion of the housing will be at a lower temperature and not letting it expand like it wants to...thus there can be a large thermally induced compressive stress that after cycling multiple times often leads to stress cracking.

Secondly, if your combustion process isn't correct, you could still have combustion reactions occurring at this late downstream region, which means that the stagnation gas temperature that is getting imposed on that section can be well above the melting temperature of the nodular cast iron that it is made from. Or, if your dilution jets are too far aft, you could have poor mixing and thus to high a gas temperature entering the turbine housing. Or, if the dilution holes are too large, your cooling air may not penetrate deep enough into the combustor exit gas flow and thus the "core" flow might be substantially hotter than the outer flow that has mixed with the dilution jets and thus you impinge a hotter core flow on the divider and melt it out....and of course, the melted cast iron goes into your turbine doing nasty things.

Finally, the extra "wetted" surface (a term meaning the fluid in the passage touches that surface, in this case hot combustor exhaust) means additional viscous flow losses and a larger heat transfer surface area. The viscous flow losses will show up as a loss in total pressure thought the housing, which means the turbine and exhaust nozzle have less pressure energy to work with. The extra heat transfer due to the increased wetted surface will be a smaller effect, but it is a loss of the flows internal energy none the less, and again there is less energy for the turbine to work with.

That is why it is preferable to use a single entry housing, but if that is all you can get, you can make it work. My first engine had a twin scroll housing and it is still intact. I profiled the divider so that it wasn't so much of a flow blockage and also so that it would have a minimal flow stagnation area...and I made sure my combustion was completed prior to the dilution holes, so that wasn't subjecting it to higher temperatures than necessary (as much as I could with atmospheric combustor testing).

I hope that helps!

Chris

[miuge]

I prefer single scroll nowadays, we did this mistake using twin scroll earlier

[witek]

Thanks for the information guys, really concise.

As a workaround however, I was thinking of fabricating some divided inlet flanges from stainless and rounding out the dividing wall to be a little more aerodynamic. This flange would be welded to the combustion chamber and would act as my connection to the housing (stainless as well).
I think the flange would act as a wear plate as stainless has a higher melting point than cast iron and hopefully save my turbine and housing.



I can live with the reduced pressure for now, as my end goal at this moment is to have a working gas turbine.

Cheers

[racket]

Hi witek

A possible solution to the overheating of the centre web is to fit a stainless steel tube across the top of the web ,it needs to be a bit larger in diameter than the web thickness , the tube having a slit in its underside just above the web , the tube can be fed with P2 air from between the flametube wall and the outer can , we need the highest static pressure possible, so if a pitot tube could be fitted into the comp housing to feed the stainless tube it would be even better.

The gases exiting the flametube will have to speed up as they try to enter the smaller scroll entry area , this will require a static pressure drop resulting in a pressure differential between the inside of the tube and the gases surrounding it , this should allow cooling air to flow out of the slit and "blanket" the web at that critical entry point .

Cheers
John

[witek]

Thanks for the tips guys. I've spent lots of time thinking of many details in this project.

John,
Has this tube solution been attempted before? I have not seen any reference to this sort of idea in my research thus far.

Going back to the option of having a stainless cross-member, would this be a viable solution? Or do you guys think that it would be over complicating things for no reason?

[Deleted]

hi

I just cut a V out of the web that splits the scroll, then ground a curved edge on the face of the web that faces the flame tube. I have not had any problems with temps etc

[racket]

Hi

I don't know of anyone who has tried the cooling blanket idea , though it has been discussed over the years , generally guys modify the dividing wall to minimise the heat "impact" on the square edged wall face or use a stainless cover piece to protect it .

Cheers
John