Nashville, TN Turbine Build

[wolfdragon]

grins and giggles how does your turbo stack up to the rules of thumb for the Compressor Inducer Vs Turbine Exducer (should be +20-25% ish)

This also is partly what I was curious about for the compressor numbers as well.

I'm sure John will show up later tonight with some nugget of wisdom that will help out quite a bit.

In the meantime, you should probably scare the neighbors some more...

[jamiep]

Here is what I am running:

cgi.ebay.com/ebaymotors/HONDA-CIVIC-D16-T3-T4-T04E-TURBOCHARGER-57-A-R-TURBO-_W0QQcmdZViewItemQQhashZitem3a5bcfb393QQitemZ250648441747QQptZMotorsQ5fCarQ5fTruckQ5fPartsQ5fAccessories

Jamie

[wolfdragon]

I was looking at one of those before I picked another

Specs after math:
77 Compressor Trim (ROT say roughly 66 is a good start)
+6.69% Compressor Inducer Area vs Turbine Exducer Area (ROT say +20 to +25% is a good start)

So that's why when you opened it up it started to surge, no place for the exhaust gasses to go

This does set the stage for some interesting data to see, try to figure out where absolute low idle is and then increase P2 incrementally up to where it surges, get pressure and temperature data over the operating spectrum.

It will be interesting to see what the operational envelope will be for a turbo that is well outside the rules of thumb.

I'm very curious now...


I am thinking that there should probably be another forum section for tabulation of such data for others to see, give the turbo wheel diameters, flametube and housing sizes, and then the operational data, probably make it a non reply section so that questions would be asked in an area not meant for a data only section

[racket]

Hi Jamie

Sounds like surge , was your P2 gauge reading bouncing around ?? ....................you can test for surge by positioning a piece of lightweight string/cotton thread so that its end is near the comp wheel , but not touching it even at max stretching as under full suction from the engine , and seeing if the thread is moving in and out in time with the chuffing sound .

It could be unstable combustion but it sounded a bit too regular for that, the frequency of surge pulses depends on the capacity of the system between comp and turb vs the flow rate of the comp , also some comp wheels can be running in surge with little indication whilst others will be "banging away like a dunny door " ........................best check it out before you do some damage to the turbo shaft and wheels , surge can wreck a turbo :-(

Cheers
John

[turbochris]

I was told if you want to test how well a turbo can handle surge, to simply beat it repeatedly with a large hammer.

[jamiep]

I fired up the engine 3 more times over the weekend...

First, I welded up a hole that the air intake was blowing directy into and opened up a few more in the tube.

Results were similar to my first run (although I did get up to 17psi on the combustor)....starts up and runs great up to about 102,000 rpms where I get surge (even tested it with the string). If I throttle it up any higher, you def get the whump, whump, whump sound. I may just have to be happy with 102k as "full throttle"

Here are some additional data points:

Idle (lowest it seemed stable)
33,539 rpms at a EGT of 1,239 F @ 2 psi combustor pressure

another area of the throttle band:
53,880 rpms at a EGT of 1,124 F @ 5 psi combustor pressure

Top end before surge:
102,540 rpms at a EGT of 980 F @ 17 psi combustor pressure


John - I know you are a numbers man, thoughts?

Also, if my tubine section can't handle all the compressor air, is it possible to make a bleed or bypass straight into the exhaust that would bleed some of that?

Jamie

[racket]

Hi Jamie

Yep , rather hot at 2psi , thats normal as comp efficiencies are low and theres lotsa bearing drag etc , reasonable at 5psi though still a bit on the high side , this is probably a "safe " idle speed .

The temp at surge is "interesting" as its relatively low , I would have expected it to be higher , mmmm, could there be some sort of "restriction " between comp and combustor , is there plenty of flow room at the junction of the air delivery tube and the combustor .

Generally , any bleeding off of air instantly raises temps and should exacerbates the surge problem as the hotter temps/lower density gases, even if of less mass flow, would have difficulty passing thru the turb stage .......................but as your engine is already running "out of tune" for whatever reason , it might be of some interest to try a small amount of bleed to see what happens , you won't need any more than a 1/2" bore bleed valve , I used a normal 3/4" plumbing gate valve when I tried bleed experiments with my TV84 turbo

Cheers
John

[racket]

Hi Jamie

Been going back over your pics and the problem could be the flametube outlet , there doesn't appear to be any "funnel" between the flametube crossection and the entry tube to the turb scroll, ....................without a funnel to guide the gasses from the relatively large crossection to the limited area of the scroll inlet , gases can "back up" and produce surge

Cheers
John

[jamiep]

I guess I could chop the flame tube about 6" shorter, then weld on a cone to transition the air into the outlet better...

I will put my latest test video up and you can see/hear what it was doing.

Jamie

[wolfdragon]

Why don't you just cut the tube that bends around from the flametube outlet to the turbine?

Then you can open up the hole size at the end of the flametube, then make the transition cone back to the size of the tube? I bet if you incorporate a slight bend into the cone you can just cut that current duct tube down a bit and weld it in place so that instead of changing the flametube, you have just made that first bit of the ducting tube bigger.

[jamiep]

wolf, that is certainly an option, but requires much more work and will affect the aesthetics.

I uploaded a video of the surge. It was bright outside, so you can't see the rpms, but I posted the data at the end of the video:

www.youtube.com/watch?v=0BWhA196cbk

Jamie

[racket]

Hi Jamie

Now thats surge :-( .......................the string certainly gives a good indication of the backflow, and combined with the P2 g/g bouncing , theres no doubts .

Cheers
John

[jamiep]

What do you think about my cutting the flame tube shorter and coning it towards the exhaust tube?

jamie

[racket]

Hi Jamie

Yep , no problems :-)


Cheers
John

[turbochris]

Sorry I couldn't stop by. Got a late start out of Iowa and I didn't get to nashville until 11:00pm and I wasn't awake anyway so I didn't know to stop driving.