I remember seeing a very good reference on various types of diffusors, with tables and correlations for all kinds of annular diffusors.

Now that I need it I can't find it anymore and I don't remember who wrote it or who published it.

Any hints on a good standard reference on annular diffusors?

ESDU publish vast amounts of this type of information:

www.esdu.com

However, the only company I know who actually designed an annular diffuser using such data sheets got into trouble because they did not understand the fluid mechanics of diffusers. They looked up a diffusion angle for their flow conditions and built their diffuser. Unfortunately, they had to add a bend or two to make it fit and added a strut or two to take load...

CFD has many failings but this type of thing is something it is good at subject to: (a) using sufficient grid resolution (coarse grid => excess diffusion => diffuser hangs on too long) and (b) being careful/sensible with RANS turbulence models which nearly all tend to separate a bit late when you push the diffusion angle.

Might want to also try http://naca.larc.nasa.gov/reports/1949/naca-rm-l9b28/ This is an oldie with limited scope, but a goodie nonetheless. I have found it extremely difficult to find good references on this subject. Looking forward to see if anybody else responds.

Japikse has written a large review on diffusor design. You can buy it from Concepts' web page. Here is a link to it:

http://www.conceptsnrec.com/education/pub_books.htm#two

As a side note to the comments made by andy and BeachComber. I think that both of you are refering to inter-turbine or inter-compressor ducts. You might want to know that there is a large proposed new EU framework 6 project devoted to this subject. The project is called AIDA (Aggressive Intermediate Duct Aerodynamics) and involves most of Europe's turbomachinery industry. AIDA hasn't been approved yet but if it is approved AIDA will provide a lot of new experimental data on this type of annular duct flows.

Interesting comments regarding AIDA: looking forward to see if this becomes funded project.

Regarding NACA reference: This is a general axial test facility with basically fan rotor (to set up swirl conditions) discharging into annular diffuser. Area increase is due to shrinking inner wall down close to zero (with respect to inlet dimensions) while holding outer wall diameter constant. Flow then goes directly to exhauster.

Regarding diffuser design book mentioned: a word of caution. For annular diffusers, there is one section devoted to the subject. We were able to extract limited useful information, and feel that it contains basic errors. One of Japikse's subsequent papers that I believe was presented a few years ago in Cleveland at NASA's Thermal and Fluid's Analysis Workshop contains attempts to correlate annular diffuser data – would suggest this as better reference, but it also seems to contain basic errors as well as what I would describe as very creative interpolation of data with large scatter.

Like I said prior, it is difficult to find good, inclusive references on this subject. Definitely hoping this EU AIDA project will advance.

Klein, Progress in Aerospace science, 1996??/

Yes, you can download also this paper from concepts' web-site. It is entitled "Correlation of Annular Diffuser Performance with Geometry, Swirl, and Blockage". You can find it in the list on this page:

http://www.conceptsnrec.com/education/pub_papers.htm

You have to register to get access to it but it is free.

I also hope that AIDA will be approved. There is a big need for this kind of data. With todays high-by-pass-ratio engines with very large fans and small high-pressure ratio cores the demands on the inter-ducts are becoming very important.

If you are interested I have a bunch of references on this type of transition-duct flows - Loughbourough Univ. have published several papers. There are also several papers from a Japanase groups as well as a few papers from other groups.

I am curious about where you believe the difficulty lies with interconnecting ducts? Certainly the correlations used in industry (at least a few years ago - I am bit out of touch these days) are of modest ability but that seems to have been more to do with the component not being rated as particularly important (i.e. no resources allocated to developing the design methods) rather than anything challenging about the flow through the component.

The geometry is simple and straightforward to predict to any degree of accuracy required short of full DNS. A good LES prediction is going to challenge the accuracy of most measurement techniques. Performing such a prediction is perhaps not yet routine but the aerospace industry can access such codes/people if the requirement is there to develop the design methods further.

If I recall correctly, the biggest problem is that the flow in a "high performance" duct is very sensitive to the inlet conditions and these conditions are generally not known in sufficient detail particularly during the preliminary design stage. Since the consequences of stalling the flow in such ducts is serious they have ended up as conservative designs. I suspect having masses of experimental data and masses of CFD predictions is not going to alter things much other than building a bit of confidence - perhaps I am getting old and cynical.

I think that you've summarized it quite well. Inter-ducts have not been researched a lot before. The reason being that in old engines these ducts were not regarded as very critical. Most of the work has actually been done at your old university (LBORO) - I think that I've seen your name on a few old papers even .Things have changed though and the AIDA project is a clear indication that industry is now focusing much more on these ducts. AIDA has 16 partners and includes all major turbomachinery companies in Europe. 5 duct-test-facilities using both turbine and comressor rigs at leading Universities across Europe will be used in AIDA.

Future super-silent ultra-high-by-pass-ratio engines will demand a new class of very aggressive (large radial shift, short lentgth & high diffusion) inter-ducts. And as you point out old designs have been very conservative due to the high costs and serious delays associated with a failing duct. Since these ducts connect engine modules they are usually not tested until very late in an engine program so any problem is bound to be very expensive.

I don't agree with you that it is an easy CFD problem - you need to predict separation from a curved wall in an annular diffuser with large radial shift. The inflow is transient, often swirling, with incoming wakes, tip-leakeage vortices and sometimes even rotating fish-tail shock structures. The outlet also has periodic pressure disturbances affecting the duct behaviour. You can even have a complex interaction between a partially separating duct and a down-stream compressor which goes into rotating stall or surge. Ducts also often have struts or even partially integrated vanes (swept and leaned with a 3D shape to improve performance).

AIDA will also address novel ways to control separation in this type of ducts using flow-control devices on the wall.

I did not say it was an easy CFD problem but that the physics involved can be simultated to a high degree of accuracy by a competent investigator using current state-of-the-art LES methods. I was referring to current inter-connecting ducts and not ones with small devices on the wall to stir up the boundary layer (reliable simulation of which is likely to be challenging since important physics will lie in the unresolved motion). Separation and curvature are mostly certainly problems for RANS turbulence models but much less so for LES.

When gas turbine engines become super-silent (non-rotating?) I suspect it will be time to retire the engineers.

You may refer:- "Characteristics of Combustor Diffusers", by A.Klein in Prog. Aerospace Sci., Vol 31, pp 171-271, 1995