Engine Dynomometer Induction fan setup?
Engine Dynomometer Induction fan setup?
(OP)
Hi guys,
I am an ME and I run a business designing and engineering engine parts for VW / Audi cars. We have two dynomometers, an engine dyno (engine out of car) and chassis dyno (full car in a room)...
Long story short is we are located in Salt Lake City, Utah- with an atmospheric pressure frequently in the region of 860 mbar. We commonly run into situations where we're wondering if the existing hardware can handle the full demands of the engine at sea level, and yet- this is a ~thousand mile drive. We're often wondering how to calibrate so to avoid running the engine into fuel system limitations and so forth.
Soooo, I'd like to build a system to raise the pressure in an air filter plenum to sea level. It's difficult to do the entire room as our chassis dynomometer is not in it's own room, rather sharing a corner of our warehouse facility and so forth. Unfortunately without moving to a larger facility this one won't happen.
It's pretty simple for me to look at some fan curves and see that obviously I need a lot of fan horsepower to bring 800 CFM or so up 60 in H2O (150mbar). It seems not too many fans may be happy with this in a single shot as well. I'm not super familiar with which fans may be appropriate for this type of duty and / or the most cost & energy effective.
So that's one topic. The other is; how to provide air pressure control in a highly transient environment. I'm looking at some data and it's fairly typical to have the mass airflow varying somewhat rapidly. The long story short is I see two methods off the top of my head: PID control it- but a large fan or fans may have a lot of inertia, making that difficult.... Another idea would be a large mechanical poppet style valve with a soft spring - and just run the fans flat out (or at a steady frequency appropriate to the engine being tested)
In any case, I'm hoping to pick your brains a little and get some practical ideas to avoid inventing the wheel here more then necessary.
I also attached a screen shot of a typical airflow meter trace during a run. Typically we would ramp the engine at a speed of approximately 200 to 500 rpm per second... So from this you can get an idea of the transients.
Here's the way one guy ran an axial fan into the engine's air intake. However, He's in Texas and I doubt he needed to raise pressure nearly like we do.

I am an ME and I run a business designing and engineering engine parts for VW / Audi cars. We have two dynomometers, an engine dyno (engine out of car) and chassis dyno (full car in a room)...
Long story short is we are located in Salt Lake City, Utah- with an atmospheric pressure frequently in the region of 860 mbar. We commonly run into situations where we're wondering if the existing hardware can handle the full demands of the engine at sea level, and yet- this is a ~thousand mile drive. We're often wondering how to calibrate so to avoid running the engine into fuel system limitations and so forth.
Soooo, I'd like to build a system to raise the pressure in an air filter plenum to sea level. It's difficult to do the entire room as our chassis dynomometer is not in it's own room, rather sharing a corner of our warehouse facility and so forth. Unfortunately without moving to a larger facility this one won't happen.
It's pretty simple for me to look at some fan curves and see that obviously I need a lot of fan horsepower to bring 800 CFM or so up 60 in H2O (150mbar). It seems not too many fans may be happy with this in a single shot as well. I'm not super familiar with which fans may be appropriate for this type of duty and / or the most cost & energy effective.
So that's one topic. The other is; how to provide air pressure control in a highly transient environment. I'm looking at some data and it's fairly typical to have the mass airflow varying somewhat rapidly. The long story short is I see two methods off the top of my head: PID control it- but a large fan or fans may have a lot of inertia, making that difficult.... Another idea would be a large mechanical poppet style valve with a soft spring - and just run the fans flat out (or at a steady frequency appropriate to the engine being tested)
In any case, I'm hoping to pick your brains a little and get some practical ideas to avoid inventing the wheel here more then necessary.
I also attached a screen shot of a typical airflow meter trace during a run. Typically we would ramp the engine at a speed of approximately 200 to 500 rpm per second... So from this you can get an idea of the transients.
Here's the way one guy ran an axial fan into the engine's air intake. However, He's in Texas and I doubt he needed to raise pressure nearly like we do.






RE: Engine Dynomometer Induction fan setup?
Off the top of my head perhaps 2 x HP-4C with 22" wheels in series on VFD's and then a large plenum to smooth airflow variences and a large spring loaded poppet valve may be the simplest (but least energy efficient?) way forward? Maybe I should design for 35" of head pressure per fan just for some safety factor and to account for pressure drop in the ducting.
Fans just as an example I understand there are a zillion manufacturers and so forth.
RE: Engine Dynomometer Induction fan setup?
RE: Engine Dynomometer Induction fan setup?
So I think I'll wait a year and slip that one through accounting separately ;)
RE: Engine Dynomometer Induction fan setup?
RE: Engine Dynomometer Induction fan setup?
As to actual fan type, either axial or centrifugal will work.
RE: Engine Dynomometer Induction fan setup?
I'm sort of thinking fast control of air pressure via a big butterfly valve driven by a servo might be the ticket. To be honest the power consumption doesn't really worry me too much. Probably looking at a monthly ~duty cycle of only a few percent. A servo (or stepper) butterfly rig would be able to respond in tenths or hundredths of a second. I'd probably start there I guess and maybe add a second, slower responding PID loop on the fans later. I don't like layering PID's all at once though, that is a lot to chase around.
We talked about a big supercharger and drive motor- I'm just not sure it'd do anything that two off the shelf fans wouldn't do and I'd have to build a frame and rig up safety guards and so forth. A big centrifugal blower would be the best bet, roots would be kind of a mess for packaging. It might be cheaper I suppose.
RE: Engine Dynomometer Induction fan setup?
Regarding the control:
You don't need to constrain yourself to reacting to changes in airflow.
The changes are completely predictable, because you control the engine's throttle.
RE: Engine Dynomometer Induction fan setup?
http://inovair.com/
Yea, the reason I'd like to have it reactive is so I don't have to spend time calibrating a baseline fan or valve curve every time I put a different motor on there... The 3rd option of course is a preprogrammed duty cycle curve for each engine and then a small amount of PI or PID control on top to dial it in for the day / car. The turbo cars too have a pretty nasty airflow gradient when the turbo spools. At the end of the day I'd be ok if pressure wavered a bit right then, as long as it's steady afterwards.
RE: Engine Dynomometer Induction fan setup?
RE: Engine Dynomometer Induction fan setup?
I'm moving a little slowly on this but...
I have 2 x fans on order now that will deliver 1200-1500 CFM @ 70" or so. Went with a pair because the startup and motor requirements are much, much less, and also a pair will hit 60" at a much lower flow rate then most large singles... Plus it was cheaper.
Anyways...
We use a lot of electronic throttle bodies around the shop. The plan at the moment is to use a pair of large throttles, PID controlled off an arduino with pressure sensor to control that.
I need to put a plenum in line, for the "weather station" for the dyno, as well as possibly the ECM (has ambient pressure sensor onboard) and so forth. This will also give me a mobile pressure chamber I can roll up right next to the car, as close to the standard airbox as possible... I was going to run large flex hose from the fans to that.
Can't decide what volume to make this.... A large plenum is slow to control - a small one will put a lot of noise into the pressure sensor... Anybody have any thoughts? I'm thinking a volume which changes over very quickly, perhaps half a second or less- to ensure we can rapidly control the pressure.
Progress though.