Flow testing with activated valve?

Post by **SchmidtMotorWorks** » Sun Jul 17, 2005 2:04 pm

Do any of you do head flow testing with a mechanism to open and close the valve at cam speed to test the influence of the pulses etc.?

(I know, it's called a dyno)

I know it wouldn't exactly duplicate the real engine but it seems like it might be useful for something.

The reason I ask, is a guy asked if I would quote making a one cylinder block and crank for him to dyno test with but the price seems like it will be more than a full block and I think the same would be true of some the other parts he will test on it like cams. It does seem like a cool way to test runners and headers though, so I was thinking maybe just a way to activate the valves on a flow bench would be a good compromise.

If I could make ten of the one cylinder blocks, the price might be worth it.

bill jones · Post by **bill jones** » Mon Jul 18, 2005 1:40 pm

-I see a couple of potential problems with that idea.
-It would be easier to do overhead cam engines but you also have the issues of how many valves would you intend to be opening during this testing?
-If you had something simple like a Subaru would you want to open just one pair of valves at a time or would you use say the stock cam and open both cylinders even tho you were testing just one cylinder at a time?
-If you had other cylinders worth of valves being opened then you'd need to have notches made in the head perch to allow those redundant valves a space to open.
-Then if you got carried away and decided to do overlap testing???
----------------------------------------------
-I think the power it would take to open the valves, installing rocker arms, the oiling problems, the noise, the jackhammer effect of the valve opening and closing would all be pretty serious issues to consider especially when all that is being transmitted into your flowbench.
------------------------------------
-And how fast do you intend to open and close the valves?
-If you were doing 10,000 rpm and big ol' heavy valves wher eyou have to have 350 to 400 closed and 1000# open pressure I can guarantee that the noise would scary as hell, same a using a spintron.
-Probably the best deal would be to use an engine block of your choice combined with spintron and connect the flow equipment to that.

Mike Rogers · Post by **Mike Rogers** » Mon Jul 18, 2005 2:39 pm

How would you even get a reading with the manometer flucuwaiting out of control from the valve opening and closing.Would you not need some kind of computer to control your settings.Or would you use something compelety diffrent.

Post by **SchmidtMotorWorks** » Mon Jul 18, 2005 4:24 pm

Good points, I wasn't thinking of four valve stuff.

I was thinking of making something like a cylinder base with a place to put a camshaft on some kind of roller bearings. Maybe make a pushrod with roller tips so that oiling wouldn't be as impotant.

Yah , I guess it might be rather loud to have in your face but maybe you could put a valve cover on it?

There would be lots of things that wouldn't be right about it like the absense of the power stroke. It just seems like there might be something of value that could be discovered this way. I was thinking that maybe you could test the intake flow for different length runners or different bowl shapes. Maybe you would even get a chance to see the shape that the airflow takes on reversion if you used a strobe light and something suspended in the air.

Ape · Post by **Ape** » Mon Jul 18, 2005 5:27 pm

actually that sounds rather interesting, if you re really interested about it search the sae papers on unsteady flowdevelopment using dopplerlasermetrie. research like the one mentioned gets conducted, but i dont know if its for a normal human being somewhat affordable. I know that some bigger technical colleges do stuff like that so there definitely is a point.
But why not testing the intake flow and bowl flow doing steady state flow tests??
anyways great idea

christian

Post by **SchmidtMotorWorks** » Mon Jul 18, 2005 5:36 pm

But why not testing the intake flow and bowl flow doing steady state flow tests??

I was told by some fuel racers that the reason that the bowls are so big on fuel heads is so that there is a large volume of presurized air/fuel mixture at the valve when it opens. I suppose there is some optimum of bowl size to either soften or sharpen the connection from the curtian area to the more duct like part of the port.

Thanks for the other references, I will look into them.

Post by **gofaster** » Mon Jul 18, 2005 6:10 pm

Mr. Schmidt, I like the ideas you bring to the forum. I hope you succeed in finding a suitable shop in my neck of the woods.
Jerry Branch is best known for his work on Harley Heads. In his book, "Branch Flowmetrics", vol. 3, 1999, he mentions a time when he did some R&D for some Japanese motorcycle companies and they let him read the papers on their tests in Japan.
Jerry wrote:
" They used a single 500 single overhead engine for the tests. In the backside of the carburetor, through the manifold, intake port,and in the combustion chamber, tiny sensors smaller than a hair were installed on the surface. Each was attached to a computer for recording the data. The gasoline was made radioactive so that it could be recorded. The information was quite surprising. From the liiquid in the float bowl til after the float valve had closed,there was only 8% atomization.Eight percent. What we see going through the bore of the carb and manifold is liquid, not gas,as we cannot see gas...."
Jerry is mostly retired now, but it's rumored that he still comes around the shop . This experiment sounded like the sort of thing you are interested in. Maybe you could talk to Jerry or one of his employees and get a little more info on this. His web site is:
http://www.branchflowmetrics.com/Heads01.htm
Jim

Ape · Post by **Ape** » Mon Jul 18, 2005 6:25 pm

Fuel heads do you mean nitro+supercharged?? sorry to sound naiv but as mentioned otherwhere im a european motorcycle guy. do you also do CFD since i read your very interesting modelling thread??

Found a nice link:

http://www.swri.edu/3pubs/brochure/d03/ ... owbenc.htm

Hhuuumm maybe one could substitute the dynamic environment with steady flow in changing testpressure according to local pistonspeed(angle oriented)??? and in doing it in increments on would get somewhat dynamic maps?

Post by **SchmidtMotorWorks** » Mon Jul 18, 2005 7:55 pm

That is a cool link! I saw something like that at a US OEM lately, at first i thought it was ajoke but these crossover ports are being made.

Interesting that you should ask about the CFD. The new intergrated mudule is just coming into being in UG nX now. I hope to have some images to show in a day or two.

It will be of the exhaust port of a nitro supercharged fuel head.

panic · Post by **panic** » Tue Jul 19, 2005 12:18 am

Smokey did this 25 years ago (tested the engine in real time while spinning off big electric motor) and found nothing useful compared to normal flow bench.

Ape · Post by **Ape** » Tue Jul 19, 2005 4:42 am

Hi

what does the abbreviation UG nx stand for??? (ultragraphix??)

yes im very excited to see the pic´s. In case of your tunnel ram, there are some really good links aboput scient. research on wall film thickness and acceleeration rates. i could post em if you want too. you would have to use babelfish since they´re mostly german, but reall informative, fueldropletsize, speeds, etc.

looking forward to seeing the pics.

cristian

PS. What computer do you run to be able to process CFD data?

Post by **SchmidtMotorWorks** » Tue Jul 19, 2005 7:29 am

NX is the new name for Unigraphics since SDRC IDEAS has been merged with us.

The new CFD stuff comes from IDEAS, it is expected to be functional any day now. I hope to try it later today.

I am interested in the links if you can post them.

Post by **SchmidtMotorWorks** » Tue Jul 19, 2005 2:48 pm

Go faster,

Thanks for the Harley link, I would really like to talk to someone that knows Harley heads. I would really like to know why there haven't been any narrow valve angle heads made for Harleys (at least none that I have seen).

Maybe you could lure Jerry to this board?

The point about "atomization.Eight percent" sounds like it may offer more return on effort. I wounder if this is the case on the kind of engines I am most involved with; carbureted SBC and BBC. Maybe this issue could be expaned on in another thread?

RRBD · Post by **RRBD** » Tue Jul 26, 2005 12:06 am

Panic, Smokey did learn alot from the Smoketron, this single machine paved the way to todays technology. Look at the Spin Trons used by cam manufacturers.

Schmidt, I would think that your pulse frequency of your flow bench should be at a rate that would be similar to piston motion, and then test with valves at common lift values.

Post by **SchmidtMotorWorks** » Tue Jul 26, 2005 10:22 am

Speaking of Smokey, has anyone read his books?