Anti-surge turbo compressor housing machining.

[NobleConcepts]

Application:Gas turbo.
I am considering an attempt to convert a comp housing from a Garrett GT2871R (472560-15) for anti-surge. The comp housing is a "GT28R" style.
I am trying to operate this unit with the following comp map points.
Flow: 20%=7, 50%=16 and 100%=38 lbs/min respectively @ a pressure ratio of 2.45. I am in a "weird" area with this engine in that I have ran a GT28RS and ran into choke conditions. Thus, I tested with the GT2871-15 and as predicted, I experienced comp surge at the lower flow regions while trying to pressurize the cylinders.
I have reviewed the compressor work of Ishida Masahiro and F. B. Fisher dealing with (MWE) Map Width Enhancement, but I can not find any technical documents describing the methods used to determine the physical placement and dimensions of the MWE groove in relationship to the comp impellar.
Can anyone direct me to info as far as groove position and dimensions relative to a twin blade comp impellar?
I have seen some diesel comp housings with the MWE grooves, but I would like to know from someone with experience if the groove position in relation to the impellar for a diesel application is the same as for a gas application.
Can I just "lift" the groove dimensions from a diesel comp housing and proportionally apply the dimensions to another housing?

 

[willeng]

NobleConcepts,

I think the problem with copying groove placement & position from another Turbo will be dependant on the comp impellor being the same, if it differs at all it won't be as efficient.
As you know the MWE groove is a slot on the sides of the compressor wheels blades. Instead of air delivery just from in front of the impellor it also can pull it from the side of it as well.
On Twin blade comp impellors i have notice that some only feed the second impellor via this method.

Others may disagree but it would seem that the placement of the MWE Groove should be in the uppermost tip area where maximum advantage of the entire blade width can be had.
If the groove is lower on the impellor it would seem to decrease efficiency.

Right or wrong--my thought's only.

 

[NobleConcepts]

willeng,

Thanks for sharing. I know that reverse engineering is not the "noble" thing to do, but I think if I see consistency in the placement relationship it may make my chance of first attempt success increase. This is a twin parallel setup, so the cost rises fast. I am testing in a engine dyno cell and I am a private operation, no financial backers.

 

[jbond]

Noble,

It sounds interesting and I regret that I have nothing to contribute. Care to share some more details about the engine you're working on?

My road car (almost finished putting the engine together) has twin GT28RS turbo's in parallel. It's a VG30DETT.

What do you think of this compressor map:
GT2871R - 743347- 1

As opposed to the one you suggested:
GT2871R - 472560 - 15

Note the above map is of the -10 config, however the compressor specs are the same and the map is the same so I posted it anyways

 

[willeng]

Just thinking, have you thought of controlling your surge problem to eliminate it before modifying your housings?

This could make a very interesting & informative thread as there are members here with good Turbo knowledge.

Warpspeed & Pat come to mind among many others.

 

[NobleConcepts]

jbond,
I have a set of the 743347-1 on the way.
The engine I am testing has a prototype split plenum on a 3.0L V6 and a camset that was specifically ground to max exhaust energy to minimize boost threshold and maintain the exhaust energy to the max RPM.
The peak Trq is approx 22lbs/min and the setup appears to be able to handle approx 40lbs/min.
I am in a flow zone that appears to be void for the GT products.
If you look at the surge line for my points 7lbs/min to 16lbs/min you can see that with the 743347-1 there is no surge "margin" at all, if any. Once again, cost X 2 is costly waste for trial and error.

 

[NobleConcepts]

willeng,
The problem is that peak Trq is at approx 22lbs/min with these cams. Trying to de-tune in this flow region is counter productive. I could decrease cylinder pressure in that region, but that is self-defeating.
I have Motec M800 that I am using and it has maxed out "outputs", I was thinking of syncing a compressor bleed via PWM in concert with wastegate PWM, but my very capable ECU is pinned out to the max for my other "toys".

 

[willeng]

NobleConcepts

I was thinking more along the lines of improvements & not de tuning as such.

Just a question about your issue,

Does your Net Flow Rate remain constant when you encounter what you believe to be Surge or are you getting a COMPLETE or Partial reversal of flow.

There a lot of misconception of the meaning of true surge & just wanted to make sure were on the same page so to speak.

 

[NobleConcepts]

willeng,
This is true, low flow, high PR compressor flow instability.
My video cam (Sony) crapped or I would have been able to video a rare event on the engine dyno.
No misconception here.

 

[willeng]

NobleConcepts,

Just wanted to make sure it wasn't Rotating Stall conditions.

The fix to rotating stall conditions & mild & true surge are one & the same--well almost.

By applying a jet--jets of air via active control to the comp impellor when approaching stall-surge conditions can & will make a marked improvement in the stalling mass flow, not to mention spool up assist.
There is more of an improvement to be had in oscillating the jet stream, 400-500Hz would be a good starting point.
Stories of another 4% is there to be found by the oscillations on top of 4-18% reduction of the stalling mass flow are not uncommon with just steady jet flow.

Air delivery from the jets will need to be in the region of 1-5% of the total compressor flow, depending on the TIP speed of the impellor.

This is not new Tech by any means & it is seldom talked about & for good reason if you want to win or are in the race to build a true antisurge Compressor, Turbine etc.
It is a fairly straight forward fix to a seemingly overwhelming issue, a few small correctly positioned holes in the housing for jet placement & big improvements are to be had.
Timing the jet stream near stall-surge condition is important as is the placement of the airstream on the tips..

There are quite a few race teams-worldwide, no names mentioned who use this technique to GREAT advantage when the rule book states specific components to be used.
Many of these systems are not found by scrutineers & are hidden inside casings etc--I personally don't like cheats, if you have to cheat to win then your not winning in my book.

 

[willeng]

Forgot to mention that these jets can be used with & without the MWE groove, jets when used with the MWE grooves are then called J-Groove & not MWE groove.

 

[NobleConcepts]

willeng,
I remember reading some docs from turbomachinery aerodynamics about the use of jets to disrupt the outer most laminar flow, interesting.

 

[j79guy]

willeng: The jet air supply being from the downstream side of the compressor, in effect increasing the mass flow across the compressor, or from an outside source? (Bottled air.) Thanks for sharing this info, I for one did not know of it.

j79 guy

 

[jbond]

To help "low-down" torque etc, would adding a supercharger help with the application? (ie twincharging).

 

[ScottB125]

By applying a jet--jets of air via active control to the comp impellor when approaching stall-surge conditions can & will make a marked improvement in the stalling mass flow, not to mention spool up assist.
There is more of an improvement to be had in oscillating the jet stream, 400-500Hz would be a good starting point.
Stories of another 4% is there to be found by the oscillations on top of 4-18% reduction of the stalling mass flow are not uncommon with just steady jet flow.

Air delivery from the jets will need to be in the region of 1-5% of the total compressor flow, depending on the TIP speed of the impellor.


Do you care to elaborate on that more? What type of jets? How many? How are they placed?

Thanks
Scott