It could be done . Cut off the flange , cut the bend down to a minimum and it would clear .

was thinking of cutting off the flange and doing just that as well....

 

Has anyone measured the pressure drop across that flange to be sure?

It look like it would be very bad for flow, but hwy not get a few readings pre and post elbow to be sure before you start hacking stuff off.

 

Haven't we already had this discussion? There is supposed to be a pressure increase - that is how we create pressure in the first place.
The neck is part of the volute. The pressure/velocity relationship between the discharge slot and the outlet is the way centrifugal compressors work.
If you made a bend with a higher velocity (which is what you are really talking about), you would just be dropping the velocity elsewhere to make pressure.

All we care about is flow, which is a function of the entire system's velocity.

 

Makes a lot of sense . But considering what MM had to say above we still couldn't be 100% sure that flow would be any different until we actually tried it .


We also care about charge air temperature and if we have to compress the air more than we need to that will increase the charge temp.
So even if flow does not change from doing this it could still be beneficial in reducing charge temps at high flow rates .

As an aside : I made the following observation when hooking a boost gauge to the charge tube (pre IC)
At whp below 320 the pressure gauge was steady at maximum flow (and around 14psi)
At 330whp+ the pressure guage started fluctuating rapidly between 14-16psi .

I suspect this fluctuation is to do with the turbo reaching it's maximum flow at that pressure .

 

But we have to convert that velocity to pressure somewhere and the best place to do it is at the volute where the adiabatic efficiency is the highest and the charge still has the greatest possible potential to lose heat along the way.

 

Ok I can see that . There must be a point that this will work against you though . EG if you blocked off the turbo and only had a 1/2" hole for the air to come through .
I still think that it is an issue for those wanting to push the limits . You don't see that much restriction as part of any design on any high hp setup .

 

A hole (restrictor) and a volute are not the same thing.

 

ALSO : A 90deg bend going off at an angle to the volute , is not part of the volute .

 

Is the diameter still expanding at a constant rate?
Yes - it is part of the volute.

 

According to this definition the volute ends at the point the discharge straightens , just before the flange .
Firstly - there is no increase in area as it turns the 90degrees in the elbow .
Secondly - the fact that the elbow deflects the air 90 degrees to the volute .
After the bend it does start to increase in area but at that point there is no change in direction so that cannot be part of the volute either .

air has to be changing direction AND expanding at the same time to get the effect of a volute .

 

That is not quite what it says, but that is wrong, anyway.

Yes, there is. Go back and measure it.


Apparently, you still don't understand it's function, even though you just repeated it a few lines up.

 

I will .

Air has to silmiltaneously expand (reducing velocity) and change direction(converting the kinetic energy into pressure) . Do you have a different explanation ?

 

And? What is the air doing in the discharge elbow?

 

I'll tell you once I get a chance to measure it .
But even if it does expand i'm sure you would agree that it is perhaps the worst volute design you have ever seen , and that it can be improved on .

 

Don't waste your time.

No nearly the worst (just look at any Saab or Volvo), but anything can be "improved". But at what cost?
I can "improve" a toilet by making it out of gold. **** will still stick to the bowl.

 

If that is as close as i'll get - i'll take it


Stuff all if i do it at the same time as my next turbo rebuild - which is the plan.

Tried teflon ?

 

Cost of Teflon Coating? Frying pans aint cheap

 

But toilet brushes are. Which is my point.

 

That is all the constructive criticism I need lol

I'll slow down on the timing a bit

 

Alright so just an update..I backed it off to ~12 degrees around peak torque and 20 degrees near redline, with an 11.0 AFR.

Is that better? Or am I still fingering explosion territory

 

That is better - if you have good fuel.
Play with the ignition timing on the dyno. If backing it down a few degrees (up to 6 or so) at the torque peak (the real torque peak, not the OE torque peak since the turbo moves it down a 1000 RPM or so) doesn't lose power then go with that.

 

Is there any logical reason I should not revert back to the internal wastegate? The guy I got my kit from welded an external wastegate on and wired the internal shut (I have not seen this to even see how well it was done I am still in Iraq). Anybody willing to do wastegate 101 lesson for me?

 

I'm really curious to see how he got an external wastegate down there

 

At the flow levels the Greddy is capable of the internal wastegate is perfectly capable of doing the job . There are however a few issues with the design that tend to cause problems for people .
1/The actuator is very weak and will not support boost at high levels at rpm over 6000 . This in itself is not an issue for the stock turbo as it gets very inefficient above about 6psi at high rpm anyway so you shouldn't be going there.
But if you upgrade the wheel this does become a limiting factor unless you upgrade the actuator or do what I do (jam the WG partially shut) .
2/The WG pivot arm bearing is just steel on cast iron and this can wear leading to the arm sticking . Never been an issue for me but others have reported problems with this . The way mine is set up the WG does not pivot very far anyway so wear is minimised .
3/I strongly suspect that many of the problems (IE boost spike) reported by various people that are blamed on the WG are actually caused by other things in the setup.

 

if you have the choice to use an external wastegate, assuming its a proven unit, use it.