Making a venturi: How inportant is having the Bernoulli shape in the throat?

[frank mccune]

     Hi All:

      I am thinking about making venturm  for engines that I wish to convert from RC to Cl.  I have had "custom" venturm made for this purpose that have a cylindrical throat with a spray bar installed.  I wish to make these venturm a bit more efficient by adding a taper above and below the spray bar.  I think that I read that these tapers should be about 7 degrees above and about 11 degrees below the spray bar. Do these numbers sound correct or are there other numbers that would be more efficient.  They sound like they are too little.

     Anybody know if the Bernoulli effect is worth adding to a venturi or is it just fluff? I have never seen a carb that did not have a constriction at the spray bar or jet in larger engines so it must be worth the extra effort.

     Comments or opinions based on fact or experiences welcomed.

                                                                                                                       Tia,

                                                                                                                       Frank McCune

[Lauri Malila]

Frank,

I think that a spraybar that goes through the venturi bore quite effectively messes up what you win with correct bernoulli shape.
I make my venturies with a small fuel post, or just a hole very little (0,2..0,3mm) after the narrowest point of venturi.
Randy recommended a 7 degree (side angle of cone vs.
center line) angle before and 11 degree after the point where fuel sprays out.
The intersection of those angles should be sharp
This works so well that I haven't had any need to experiment more.

Lauri

[Bill Johnson]

Frank, I believe there is no absolute correct "bernoulli shape". I've been working on engines from small 2 stroke up to P&W JT9s my entire professional life and seen just about every kind on inlet imaginable. The best are designed then tested on a flow bench. They tend to be application and environment-specific.

The Bernoulli effect is very real. You get it just by adding a spraybar to a straight tube because you have effectively reduced the intake diameter at that point, fluid (air) velocity will increase thereby causing pressure to decrease.

Increasing the area before and after the point fuel is drawn into the intake helps increase the air mass entering the engine. In particular, the increased area after the spraybar causes velocity to decrease, increasing pressure just befand improving flow through the engine.  The larger diameter before the spraybar allows greater air mass into the engine. Enya has recently made a good improvement in this area by introducing swirl vanes in the venturi. This increases the air mass entering the venturi improving power.

Another thing to consider is the effective area at the tightest point. It's a balancing act because the smaller the area, the greater the pressure drop which improves fuel draw. On the other hand, you are also limiting the amount of air entering the engine, thereby limiting power.

As to the angle intersection, I would tend to favor a smoother transition. Smooth subsonic airflow depends on smooth transitions. Sharper transitions can lead to flow separation from the boundary layer which is very inefficient whether it's on a flight surface or a venturi.

Anyway, that's my rambling on the subject. I'm sure some of the more knowledgeable guys here will chime in with better information for your project. Good luck! I hope to be making my own venturiis soon as well.  

[RandySmith]

Hi Frank

On venturis The best is to make roughly what Lauri has posted a picture of, with or without the small fuel post, I have several decades of using and testing almost everything venturies/resrtictors. Given the exact same area a true venturie will outperform a restrictor, it gives more power, it seems to prime much easier, and transitions are better.
I use a flared bellshaped mount at the top, going to about a 7 degree angle, then the transition point it changes sharply to a 11to12 degree angle. The change needs to be sharp, and not smooth, this is the point the fuel feed should be, with the hole just below the angle, I try to make the top of the small feed hole just touch the angle, or if you use a fuel post put it there too. We sometimes used R/C needles as fuel post that screwed in for engines that did not have the extra aluminum either front or rear of the venturie to hold the spray bar, or just made them. the OS 2A works pretty well for that. I have used many of them in OS and Evo 36 engines threaded into the venturie itself.
Do not drill the venturies out to change them without cutting the angles back into them, this will ruin them. And do not make the feed hole any larger than the hole in the spraybar.

Randy

[Tim Wescott]

This is restating to some extent: just making a straight tube and bunging in a spraybar will, to some extent, make a "venturi".  The important part is that the area is necking down (which happens in the spraybar case because someone put this honkin' big tube crossways through the throat of the thing).

I can't comment on the "true" venturi shape that Lauri and Randy are advocating, except that it looks hard to machine, so it must be good.

[Motorman]

The ends can be square and the ID straight wall and it will still work just fine at the sizes we use for stunt with a Tiger NVA across the middle so, anything you do beyond that is gonna' work.

I have a 1/4" square tool bit that I ground a concave shape and enough relief on the end and side to use it as a form tool inside the hole. You can change the angle to get different parabolic shapes and sizes.


MM

[Steve Helmick]

I've made a lot of "venturi" straight through with the chosen drill size, and carved the bell mouth with just a #11 Xacto blade, after roughing it in with a countersink. I was told it was a bad idea to even chamfer the bottom of the bore, so I just give it a minimal edge break, either with the c'sink or #11. They've always worked quite well. 

While I have made them from aluminum, I much prefer Delrin, and I mean real Delrin, not the stuff they say is the same. It does not machine the same, and generally sucks, IMO. If you want to make a few, a 10' length of real Delrin should be well under $20, so you'll have a lifetime supply. I think I get 5/8".

My process is to face, center drill, and double drill the bore, then turn the spigot, and a skim cut on the cylinder OD, all in one operation. If the spigot is the correct size (this is the most likely machining malfunction!), then I pull the piece out and bandsaw it off with a bit of excess to face off (I don't have a parting tool). Then I reverse it in the lathe, face it to the length I want and do the bell mouth. Finding the right rpm and angle to use the #11 is interesting. Not too fast, and more angle than you'd think.    Steve

[Serge_Krauss]

I'm not an engine guy, but here are a couple thoughts to mull over. It's possible that you may want turbulent flow to mix the charge better as it enters the case and then ports. In 100cc kart engines, we found that polished transfer roofs worked noticably less well than rough ones in getting the charge into the combustion chamber (maybe it just allowed it to burn better or cut exhaust losses). Also, turbulators on wings allow flow to conform better, really changing HLG flight characteristics. So, might it be possible that the sharp transitions mix the charge better? 'just wondering...

SK

[Bill Johnson]

I'm not an engine guy, but here are a couple thoughts to mull over. It's possible that you may want turbulent flow to mix the charge better as it enters the case and then ports. In 100cc kart engines, we found that polished transfer roofs worked noticably less well than rough ones in getting the charge into the combustion chamber (maybe it just allowed it to burn better or cut exhaust losses). Also, turbulators on wings allow flow to conform better, really changing HLG flight characteristics. So, might it be possible that the sharp transitions mix the charge better? 'just wondering...

SK

I think you have to have good fuel atomization but you also need smooth airfow into the intake to get the maximum air/fuel charge into the cylinder. The 7/11 degree venturi works fine because the total agular change is 18 degrees, right at/below the critical angle of attack for a typical subsonic flow surface. Any more then that and you need to energize the boundary layer like Enya does woth their new "swirl" venturi, the same effect you get with  vortex generators.

[Brett Buck]

I'm not an engine guy, but here are a couple thoughts to mull over. It's possible that you may want turbulent flow to mix the charge better as it enters the case and then ports. In 100cc kart engines, we found that polished transfer roofs worked noticably less well than rough ones in getting the charge into the combustion chamber (maybe it just allowed it to burn better or cut exhaust losses). Also, turbulators on wings allow flow to conform better, really changing HLG flight characteristics. So, might it be possible that the sharp transitions mix the charge better? 'just wondering...

  For stunt, it's actually critically important, far more important than the volumetric efficiency or flow per unit draw. Not changing flow characteristics with changes to the load or the input air direction/velocity. This is what leads to the unexpected phase changes/apparent mixture changes. Since you don't really care how efficient it is and you can get arbitrary amounts of power almost no matter what you do to the venturi, my opinion and experience has been that almost anything you can do to improve mixing or make it insensitive to maneuvering, the better. That's why you want diffusers on the input, and, if there was a way to do it, you might want a large plenum as an airbox. 

   One of the reasons the ST46 runs so smoothly is that the internal shape of the venturi is *not* smooth, making the otherwise very undesirable "flush fuel inlet" work OK. In the stock venturi, the fuel comes in through 6 holes that reside in the lee of a sharp edge, where the flow is almost certainly very turbulent and relatively low velocity.

   Note that the fuel suction is only one side of the equation - it also matters how hard it is to suck the fuel through the plumbing. Completely conventional plumbing has to flow at most about 1.5 ounce/minute on average (8 ounces for a pattern), that doesn't mean it flows freely enough to keep it from unexpectedly going lean, because the required peak flow is much higher than the average. This is the problem with 4-strokes running conventional tanks and thick fuel, but it also happens on the larger 2-strokes.

   There, that ought to get the spirited discussion started again.


   Brett

[Chris Wilson]

I............ we found that polished transfer roofs worked noticably less well than rough ones............
SK

Yep, with the rich mixtures and large droplets size used in model engines it is easy to promote condensation inside of any port and polishing simply gives a "wetter" surface for the fuel to stick to.

I think that the venturi is only part of the solution, but a simple restricter or through style of spray bar can be used to achieve a proper Bernoulli gradual  taper if you relieve the venturi around the intrusion of the bar like they do with some of the Stalker engines.

I believe that the engineering to achieve the correct amount of 'relieving' is very precise though and hardly worth the trouble.