MRT-02 new turbine project (ideas and suggestions?)

[miuge]

Hi John,

That's pretty much what I was thinking about tip height vs. PR design, but when it comes to choosing a matching wheel in a gas turbine things become more complex...

Now if I say I'm looking for the best peak power, should I choose X846 instead? There's even a X844 wheel with a 6.60mm tip height... Is it too much, in to out ratio being 2.75:1 ?

Is there any gain using the 129/112mm turbine wheel instead of smaller 125/106mm?
Seems like they're both made of same alloy, bigger wheel manufacturer has been in turbo business since 2008 and another since 2002. Both look OK in the pictures. No major difference in price neither.

[finiteparts]

Hi Miuge and John,

I just thought I should add a correction to this...

"The bigger tip height sorta indicates a lower pressure ratio design as they try to design the comps for a roughly consistent flow velocity into and out of the wheel , the increased flow area at the exducer allows the lesser density air to flow out whilst maintaining velocity and mass flow..."

The exducer tip height, referred to as the "b" height, is used to control the impeller discharge relative Mach number. By making the b height taller, you reduce the relative flow speed at the impeller discharge, W2, and this increases the amount of diffusion that the relative flow experiences as it travels through the impeller. Reduced relative flow speeds, W2, in the passages mean that you have less scrubbing losses in the passage and thus a more efficient transfer of energy through the impeller.

Due to exducer blade backsweep and discharge flow slip, the reduced meridional discharge velocity, Cm2, causes the absolute tangential velocity, Cu2,to increase...this might seem odd, but remember that the exducer tip velocity, U2, is fixed by the radius and rotational speed. So when the meridional velocity, Cm2, increase, the relative discharge vector, W2, grows and thus the relative tangential velocity, Wu2, grows. When you subtract the relative tangential speed, Wu2, from the tip speed, U2, you are left with a smaller absolute tangential velocity, Cu2.

Now for the big point...assuming that there is no swirl at the compressor inlet, then the enthalpy rise is given by the tip speed, U2, multiplied by the absolute tangential velocity, Cu2. So, increasing the b-height increases the absolute tangential velocity component, which increases the enthalpy rise across the impeller and thus increases the pressure ratio.

So if a larger b-height gives a larger pressure ratio and increases the diffusion ratio across the stage, why wouldn't large b-heights be the way to go. The reason is that just like any diffuser, there is a limit to the amount of diffusion that can be achieved before you get large scale flow separations. Different sources claim different limits for the diffusion ratio (W1s/W2), but a common value is around 1.8. This is just another design trade that the impeller designer must navigate.

I know that is a wordy explanation, but hopefully you get the idea that increasing the exducer blade height actually increases the pressure ratio and reduces the internal passage scrubbing losses, well until the flow diffusion gets too large and the impeller passages experience increasing flow separations.

Good luck!

Chris

[racket]

Hi Miuge

I'd opt for the larger turb wheel as the 112 mm exducer provides a bit more flow area and would be a good match for those ~106mm comp inducers whilst having the potential for an even slightly larger inducer.

I wouldn't go with the 6.6 mm X844 wheel , thats getting just a bit too "tight" , probably OK for turbocharger pumping 70+psi boost, but because comp efficiency starts to drop off at such high pressures they're probably not the best for us.


I like the X846 wheel, I have one sitting in my cupboard near the computer just to look at and get inspired , and with the in/out ratio of ~2:1 its a reasonable "compromise" wheel and would be a nice match for the 129/112 turb wheel .


The 106 mm exducer on Anders turb wheel has been clipped so flow might be getting some way towards the 112 mm exducer though at the expense of some gas deflection and power production

Cheers
John

[miuge]

Thanks for your answers Chris and John! Things getting more clear now Might need to order some parts soon...

[Chuks]

Dec 29, 2016 7:29:34 GMT -5 miuge said:

Thanks for your answers Chris and John! Things getting more clear now Might need to order some parts soon...

wish you successful build, its gonna be large

[gtbph]

Hi Chris,

Are you sure, is it really possible to draw that conclusion based on the tip height alone? Couldn't a higher tip height also mean the wheel has a higher flow? If it had a higher PR as you explained, the outflow angle would be more shallow than usual. This would mean the air travels a longer distance in the diffuser, and the losses would increase. Because of this I would have guessed it is more likely that most wheels are designed for roughly the same ideal outflow angle, and this one has a higher flow.
If only we could compare the inducer blade angles...

Hope everybody had a nice Christmas,
and yes, impressively large build indeed: Happy building, Miuge,

Alain

[finiteparts]

Hi Alain,

The discussion does assume similar mass flows, which is a very valid assumption. It also assumes that the inducers are identical, which may or may not be the case for the actual impellers in question, but for the sake of discussion on the impact of b-height it was assumed.

The controlling area in centrifugal compressors is the inducer throat, not the exducer area. Expanding that further, in a compressor system, the diffuser throat can also be a controlling area. The fact that the relative flow through the impeller diffuses should indicate to you that the exducer area does not "control" the mass flow rate.

More specifically , in reference to using it in a turbine engine, as stated by Frank Whittle, "What flows in is determined by what can get out." So, if the two compressors were swapped out in the same engine, the larger b-height impeller would produce a higher pressure ratio, for the same back pressure.

As an example of this, look at this paper from IHI...

www.ihi.co.jp/var/ezwebin_site/storage/original/application/9e53d8849d72ff65891867ea69e338a8.pdf

If you look at Figure 3, you can see that as blade height increases for a fixed backsweep angle, you get an increasing impeller pressure ratio.



I hope that helps to clear things up.

Chris

[gtbph]

Hi Chris,

Yes, I completely agree with your reasoning; if the mass flows are identical, the PR is higher for the wheel with the increased b-height. This is a very interesting and surprising fact, I was not aware of that before you posted it here.
I just wanted to point out that this entails a shallower exducer flow angle, which would probably produce greater losses, such that in the end I think the wheels might have different mass flows.

Alain

[miuge]

Hi again,

One thing I completely missed earlier was that larger wheel has a 11 blade design and smaller one is 13 blade, what's the difference? Another thing I have no clue about...  
Anyway if I was about to go with the bigger wheel, this would bring it even closer to TV94 wheel, which is apparently a 11 blade too.

[finiteparts]

Hi miuge,

Increasing the number of blades does several different things that are not necessary complimentary.

The first thing that increasing the blade count does is it reduces the amount of slip at the discharge plane. In case you are not familiar with it, slip is a deviation of the flow from the trailing edge metal angle. Slip is a loss to the energy imparted to the flow and as such, it is usually a good thing to reduce. Slip occurs because there are cross passage pressure gradients in the impeller that impart a turn to the flow as it leaves the impeller, so it does not follow the blade angles. Increasing the number of blades reduces the cross-passage pressure gradients (what is termed aerodynamic loading) and thus reduces the amount of turn that the discharge flow experiences. It is often said that the flow is better "guided" by more blades...so in this respect, more blades are a good thing.

The second thing that having more blades does is to reduce the flow through the impeller because there is more metal there blocking the area. This is why you have splitter blades. There is a minimal area that the flow sees as it is going through the inducer which sets the upper flow limit due to choked flow at this throat. So if you can have a small number of main vanes, you can have a larger inducer throat area so that you get as large of a choked flow rate as possible. But, because of the slip issue, you need to add more blades so that you don't get large slip angles or even worse, large scale flow separation in the impeller...so you add splitter blades to reduce the aerodynamic loading on the vanes down by the exducer were the relative passage velocities are the lowest.

In conjunction with this, each blade forms boundary layers which become an aerodynamic blockage, further reducing the flow area. The skin friction reduces the amount of energy that gets transmitted from the rotor to the airflow...so this is a downside for having too many blades.

Finally, the number of blades changes the sound energy transmitted from turbocharger. Noise, vibration and harshness (NVH) are a huge deal for the auto industry and there is a ton of research spent on making sure the operator can't hear or feel the turbo...I know that is blasphemy to most of us here, but that is how it is. Ok, so increasing the number of blades raises the blade passing frequency noise and depending on the turbos rpm, this may help get the emitted noise to a frequency that is inaudible to the human ear. Now I suspect that no one here cares much about this...we love the noise!

So there is a balance were increasing blade count reduces the blade loading, gives better flow guidance and thus increases the impeller efficiency, hopefully more than the added flow blockage and boundary layers reduce the efficiency, leaving you a net gain in efficiency. But, for the home builder, we really have no way of knowing these efficiencies other than the published maps. If you have no maps, you are really just guessing.

Did that help?

Chris

[racket]

Yep , its an interesting Paper that says a lot, but really doesn't tell us much .

From what I can gather the wheel is ~140 mm exducer with probably ~102 mm inducer , so ~52 Trim and their in/out area ratio is ~2.2 :1 for its 5:1 PR potential with ~2.2 lbs/sec .

Its a shame they didn't include a compressor map rather than just "dimensionless" information which for this non engineer makes life difficult :-(

[stoffe64]

Jan 4, 2016 20:09:13 GMT -5 madpatty said:

Hi Miuge,

Just for a reference if in case you need any.

m.ebay.com/itm?itemId=251420411832

I have bought one and its in perfect condition.

But take care they are reverse rotation ones.

Cheers,
Patty

What partnumber was it on your turbo patty?
Cheers /stephan

[madpatty]

Jan 5, 2017 17:19:03 GMT -5 stoffe64 said:

Jan 4, 2016 20:09:13 GMT -5 madpatty said:

Hi Miuge,

Just for a reference if in case you need any.

m.ebay.com/itm?itemId=251420411832

I have bought one and its in perfect condition.

But take care they are reverse rotation ones.

Cheers,
Patty

What partnumber was it on your turbo patty?
Cheers /stephan

Hi.
Sorry tht turbo is no longer in my possession.
I am afraid if i correctly remember the number but best guess would be:- 3594122 as far as I remember.

Cheers.

[stoffe64]

Did you buy a left rotating turbo patty?,that partnumber was for HX80 ht

[madpatty]

Jan 6, 2017 3:57:11 GMT -5 stoffe64 said:

Did you buy a left rotating turbo patty?,that partnumber was for HX80 ht

Yea that was a left rotating turbo.

HT80.

Cheers.