Airflow direction for cooling acoustic engine enclosures

[ehestnes]

We have engines installed in acoustic enclosures. The noise inside is up to 120 dBA and we reduce this to 82dBA (Average)
The rooms are cooled by a cooling fan. For our application we require approximately 1.7 m^3/s airflow, so it is significant. (ducting etc is kept large enough to have max air speed of 9 m/s.).
Engine is cooled by remote radiators and exhaust is routed out of the room. So all we are cooling is the heat generated from the high surface temperature of the engine and exhaust pipes.

My question is;
Blow the air into the room, or suck the air out of the room?
Our inlet and outlet ducts are typically placed diagonally opposite with the inlet duct low and the outlet duct high. We can swap the fan around on most of the applications.

I am personally for blowing the air in since I think we achieve more turbulence and less chance for deadspots.
We also achieve a slight overpressure which helps the Natural aspirated engines.
One of the arguments I meet is that the air forced through cracks and gaps will create more noise being pushed out, than if the air is sucked in.

I am sure there is lots oif views on this tropic and would appreciate your feedback.

Thanks

Even

 

[Dean78au]

All of our enclosures (over 30) that contain either turbines or piston engines have air forced in. I am yet to come across an enclosure where air is sucked out. If your enclosure is well designed there should be minimal leakage through cracks etc. If these cracks exist then noise from the engine can escape in a similar manner to air and drowns out the minimal noise from air being forced through the cracks.