CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   non-dimensional analysis in Fluent (http://www.cfd-online.com/Forums/fluent/37788-non-dimensional-analysis-fluent.html)

Endee September 4, 2005 11:44

non-dimensional analysis in Fluent
 
Hi, can someone please tell how to use non-dimensioanl analysis in Fluent. Like I have a known Re of 1e5 and i want to simulate this over a circular cylinder in 3D. Cheers, Endee

Masood September 4, 2005 13:51

Re: non-dimensional analysis in Fluent
 
hi Reynolds No. itself is dimesionless but when u apply it there should be any geometry. and u have to define it(geometry). any way in FLUENT i think u can simlate the original geometry itself and no need fr dimensionless calculation or scaling up or down to calculate results of required geometry.


Endee September 5, 2005 06:42

Re: non-dimensional analysis in Fluent
 
hi, thanks, but can you imagine the geometry itself cant be implemented straightaway. Re=rho*vel*Dia/Meu now if rho=1.225, meu=1.7894e-5, Re=1e5 this gives for vel calculation... vel=1.7894/(1.225*dia)

a) if your geometry is of dia 1 meter then vel=1.46 m/sec (is it logical at the Re)... b) if your geometri is of dia .0001 m then vel with respect to the same Re is 1460m/sec

What do you say?

Masood September 5, 2005 14:41

Re: non-dimensional analysis in Fluent
 
hi endee i can explain it like this. in simulations u can simulate logically bizarre situations which can be unrealistic. like u can simulate load distribution of urself as holding a 1000 ton boulder, it is unrealistic but can be simulated. or holding a feather in one hand and claculate load distribution. got my point....

u have to select a reasonable size fr ur problem. dimensionally u can reduce the size to very small levels but it will not be relaistic.

i hope i make myself clear.

Jason September 6, 2005 08:39

Re: non-dimensional analysis in Fluent
 
You have two unknowns (size and airspeed) so you need to match Reynold's and Mach number. Use Mach number to calculate your airspeed, and then Reynold's to calculate size. If Mach number isn't known, then you're going to have to infer if the original model was subsonic incompressible (Mach > 0.2), subsonic compressible (0.2 < Mach < 0.8), transonic (0.8 < Mach < 1.0) or supersonic (1.0 < Mach). Once you pick which Mach range you should be in, then you can calculate the range of sizes and airspeeds that would fit in this range. Then pick a combination that makes sense for your model.

Hope this helps, and good luck, Jason

Endee September 6, 2005 14:32

Re: non-dimensional analysis in Fluent
 
Dear Jason, Thanks, Thats seems much logical.. Just one relative query. If for only known Re No (1e5) and aspect ratio of cylinder (1-4), as Masood has mentioned, if i take some ratio, which is sort of illogical (like may be if i take the cylinder dia as .001m then the velocity becomes of the order of say two thousands meter/sec, will the result different if i assume dia as 1 meter with corresponding velocity of may be 2 m/sec. Because the Re is same for both cases. And that was my actual question. sincere regards, Endee

Jason September 7, 2005 08:17

Re: non-dimensional analysis in Fluent
 
Like I said, you have to look at your Mach number. Assuming standard atmospheric conditions, an airspeed of 2m/s is equivalent to Mach .0059, while an airpseed of 2000m/s is equivalent to Mach 5.9! That's going from highly incompressible subsonic to hypersonic! You need to try and match both the Mach number and the Reynolds number.

Mach = V/(Gamma*R*T)^.5 RNL = Rho*V*D/mu

So figure out what Mach range you should be in:

incompressible subsonic -> Mach < 0.2 compressible subsonic -> 0.2 < Mach < 0.7 or 0.8 transonic -> 0.7 < Mach < 1.0 supersonic -> 1.0 < Mach (there's also hypersonic, which is Mach greater than 3.0 or 5.0 depending on whom you talk to, but I seriously doubt you're up that high)

In the incompressible subsonic range, you aren't very sensitive to Mach number, so match your RNL (making sure that the Velocity you choose still falls in the incompressible subsonic range). In the compressible subsonic range you are somewhat sensitive to Mach number, so you should try and find out what Mach the test was done at, but you can usually make educated guesses based on the information in the test report (in newer reports, especially gov't funded tests, this information will usually be listed, but in older reports and in some university reports it won't be listed). In transonic you are very sensitive to Mach (but if you were looking at transonic test data, then the Mach number would be listed). In supersonic a cylinder isn't too sensitive to Mach (assuming you stay between Mach 1.0 and Mach 2.0 at least, higher than that I couldn't tell you, but once again they would list the Mach number for supersonic testing).

So like I said, figure out what Mach range you are in. Unless you know exactly what Mach you should be at, this will help you narrow down what size your model is. So pick your Mach range and use that to calculate a velocity, and use the velocity and the known RNL to calculate the diameter. I'll assume you're in the incompressible subsonic range (because from the way you were talking you haven't found any information on Mach number) with air at standard atmospheric conditions, so your Mach range is from 0 to 0.2 and your RNL is 1e5 (which you gave earlier).

At Mach = 0, V = 0, and D = infinity

At Mach = 0.2: V = Mach*(Gamma*R*T)^.5 = 0.2*(1.4*287*288)^.5 = 68.05m/s D = RNL*mu/Rho/V = 1e5*1.79e-5/1.225/68.05 = .021m

This give you a range to work in and you can pick a diameter that is somewhere between infinity and .021m. So pick something that seems reasonable. I would think a .05m Diameter would be reasonable for a test. Using the Diameter I just picked and the known RNL, I calculate a velocity of 29.21m/s, and double checking this results in a Mach of .086, which falls in the incompressible subsonic range.

If you're in the compressible subsonic or transonic region, you can bound your problem like I did above. Unless you're given the Mach number directly, it won't give you an exact answer, but it'll give you a range to work in. From there you can make educated guesses as to size etc. based on information in the test reports (whether it's estimating cylinder diameter from pictures or estimating velocities based on wind tunnel parameters).

You need to pay attention to the Mach number. During WWII, compressibility effects weren't entirely understood by aircraft designers. When a plane would go into a steep dive, they would lose all control of the airplane due to these compressibility effects. They were getting "Mach stall" on the wings, which is a transonic effect. So as you can see, not paying attention to your Mach range can completely change the problem that you're working on.

Hope this clarifies things.

Good luck, Jason

Masood September 7, 2005 14:24

Re: non-dimensional analysis in Fluent
 
hi Jason made the point very clear. as i can say now is get to know ur problem well and the purpose of test or simulation. then u can narrow down the variables ranges to work in.

good luck m8


Endee September 7, 2005 16:16

Re: non-dimensional analysis in Fluent
 
Thats lovely, Thanks a lot, cheers Endee


All times are GMT -4. The time now is 22:11.