Venting Analysis 1

[franck]

Hi all,

I have to perform a venting analysis to determine how many holes I have to introduce in my waveguide assembly so that there is not a big pressure differential when the assembly is going through vacuum (it is space application).

I have one article dealing with Venting Analysis, 'Venting of Space Shuttle Payloads' by Mironer and Regan but unfortunately the differential equations they provide does not make lots of sense and it cannot be numerically integrated since really unstable.

I am wondering if someone else has already performed a venting analysis and what was his reference.

Franck

 

[iainuts]

A venting analysis is often only slightly different from any other flow analysis in that the pressure drops may be very high, and somewhere along the pipe, a choked condition may exist (I'm assuming this is a compressible gas flow analysis you need to do).

I take it one issue is the varying vent pipe outlet pressure. The outlet pressure drops as altitude increases, and this pressure drop will happen quickly. The second issue is the possibility that the outlet pressure drops to the point that a shock wave forms at the vent outlet. If for example, the outlet pressure is a perfect vacuum, then virtually any significant flow will result in a shock forming at the outlet.

Anderson Greenwood manufactures relief valves and came up with their own analysis where there is a shock wave created at the pipe outlet. So that paper is one possibility, and I can send you a copy if you'd like, or call AG directly. Still, I wouldn't swear by it since I've also done analysis using programs that come up with considerably different results.

I have to believe there are numerous packaged programs available, though they may have a problem with the outlet conditions. I've often created my own itterative programs for something like this using Excel, and just your typical equations for compressible gas flow straight from Crane Paper #410.

 

[franck]

Iainuts, you are right the external pressure is dropping really quickly and in a non linear way during the ascent of the rocket. This has to be taken into account when solving the differential system. Moreover, as you pointed out as soon as the external pressure becomes lower than 0.528 times the internal pressure (for air) your flow gets choked. This is why there should be two sets of differential equations. Finally, the biggest unknown is obviously the coefficient of discharge. If it was different when you run your different programs, it could have led you to different results.

By any chance do you have the name of some programs that you used?

 

[sailoday28]

Can you describe the flow/geometry configuration?
In addition boundary/initial conditions--for example is the intitial pressure atmospheric / temp20C and then linerally (or otherwise) the outside pressure is brought to space/vacuum conditions over some finite time., etc.

If initially atmospheric, then perfect/ideal gas conditions are a good assumption.

 

[franck]

The geometry of my structure to be vented is really simple, a waveguide is a kind of rectangular beam with an enclosed volume and some circular vent holes (diameter and number to be defined). Your starting conditions are correct: close to 20C and atmospheric pressure. After that you decrease the pressure from atmospheric pressure to vacuum in approximately 70 secondes.

 

[iainuts]

The programs I've used are proprietary, sorry.

Why are you using dif-eq's? Why can't you use algebraic formulas and use numerous iterations? I have to assume you know outlet pressure (ie: atmospheric pressure) as a function of time, so doing iterative time steps on a spread sheet and ensuring your analysis is conservative should be sufficient to ensure success.

On vent lines where a tube is open to atmosphere, I've only ever assumed a resistance coefficient, not a discharge coefficient. The resistance coefficient given by Crane is 1 if memory serves.

One other thought, is the vent line open with a single opening, or is it open to a T? If it's open with a single opening there is a small thrust created. This has to be reacted by the tube supports, and may even create thrust loads which have to be corrected for if they are high enough.

 

[sailoday28]

The simplest modeling might consider:
1)The internal fixed volume of the wave guide to be adiabatic. Therefore in that volume pressure is related to specific volume as PV^k =constant or p/(m^k)=constant.
2) Mass rate of flow is simply orifice with a Cd or piping with a resistance. The flow is therefore dependent upon the upstream pressure of the wave guide, back pressure (if not choked) and flow resistance.
Coupling 1 and 2 together will result in a differential equation which can be numerically integrated.
I am curious if any of the above is included in the paper authored by Mironer and Regan?

 

[franck]

Hi Sailoday, if you want a copy of that paper I can send a copy to you.

 

[sailoday28]

Franck
I'd appreciate the paper.

sailoday28@yahoo.com