Intake Runner - Interior Finish

[markocosic]

Hey all,

Is there a consensus on the most appropriate interior finish for intake system runners?

We have fabriacted ours based around part of the bike's throttle body (chopped down and welded to alloy tube) to avoid machining injector mountings etc, and now need to blend/finish the insides.

On the bike, the runners were polished - but had throttles inside them so they would have to be to seal. Every other vehicle I've seen has had either rough as-cast or 'sand-blast' finish to the intake system internals.

1st year fluids labs (deliberately 'tripping' airflows to make them stick better, or looking at rough vs smooth surfaces) say that rough tends to deal with bends and the likes better, but there are no bends in our intake runners and the only diameter/section changes are narrower rather than wider, so I can't see this mattering too much.

In the absecnce of knowing any better (and I'm not an aero/fluids/powertrain type) I'd flapwheel to an 80-grit finish and snadblast as a finish - but if there are any texts I could be looking at (or your experimental findings) I'd like to know more!

Cheers,

[markocosic]

Hey all,

Is there a consensus on the most appropriate interior finish for intake system runners?

We have fabriacted ours based around part of the bike's throttle body (chopped down and welded to alloy tube) to avoid machining injector mountings etc, and now need to blend/finish the insides.

On the bike, the runners were polished - but had throttles inside them so they would have to be to seal. Every other vehicle I've seen has had either rough as-cast or 'sand-blast' finish to the intake system internals.

1st year fluids labs (deliberately 'tripping' airflows to make them stick better, or looking at rough vs smooth surfaces) say that rough tends to deal with bends and the likes better, but there are no bends in our intake runners and the only diameter/section changes are narrower rather than wider, so I can't see this mattering too much.

In the absecnce of knowing any better (and I'm not an aero/fluids/powertrain type) I'd flapwheel to an 80-grit finish and snadblast as a finish - but if there are any texts I could be looking at (or your experimental findings) I'd like to know more!

Cheers,

[KU_Racing]

My stance on this topic is the smoother, the better. Im sure other people will agree or disagree.

[Wesley]

Almost every stock system will have a cast finish - the reason behind this being generating a smooth finish by post-casting machining would be way too intensive for the gains you see. It would cost a auto manufacuter a fortune to polish cylinder head and intake manifold surfaces, when the gain will be minimal.

That explains stock finishes, but what is best for a race car? From an aerodynamic standpoint, surface roughness would create a larger boundary layer than a smooth surface, so in that regard, smoother is better, but an engine is not strictly an aerodynamic machine.

There is some theory on rougher surface being conducive for keeping fuel in suspension by inducing flow turbulence - fairly important in a carbuerated/TPI application, but not so much in multi-port fuel injection. It is my understanding that about an 80 grit finish is a good balance for a carbuerated application. However, with multi-port fuel injection, mirror finish will be close to optimum.

However, you also should weigh the benefits of a mirror finish versus that of 80 grit or so. How long will it take to make it mirror, and is it worth that extra fifteenth of a horsepower to do it?

[VFR750R]

I've heard both ways, but I like smooth. My comment isn't exactly on subject but extrude hone, where the diamond putty is pushed through ports or intake runners is almost always a gain in performance. Now, it's doing alot more then smoothing walls, but obviously those casting imperfections and edges which are small compared to the cross section are significant, so the smoother the better sounds good to me.

As already mentioned roughness can be used to help flow around a short turn, perhaps, but extensive testing would need to be done to determine depth and spacing of 'trips'. These trips are not likely to be formed with 80 grit either.


When it doubt, smooth it out.

[NetKev92]

Like most things in engineering, it probably depends.

If the runners are very direct, smooth should be good like the smooth trumpets of IRTBs on most I-4 racing bike engines. If you have a lot of piping and turns, you have a higher risk of separation inside the tubing, causing drag which some roughness might prevent. Like airplanes though, unless there is another aerodynamic problem already present, roughness isn't typically necessary. Vortex generators and boundary layer trips can be very useful, but they are often considered a patch.

[Brian Evans]

The consensus from people who do this for a living (port high HP cylinder heads, that is) is the as-ground finish, about 80 grit, is best. Keeps the fluid in suspension, or some such thing. Mirror finish, or even flap wheel smooth finishes, are for show, not go...

You see this on a dyno, not on a flow bench, or so I am told.

Brian

[Marshall Grice]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">The consensus from people who do this for a living (port high HP cylinder heads, that is) is the as-ground finish, about 80 grit, is best. </div></BLOCKQUOTE>

VFR750 does this for a living, and he does not agree with your consensus.

[Wesley]

I bet most professional porters haven't reached a consensus, and they still argue over which is better, and continue to do it the way they know best.

I know a few porters who know pretty much nothing of engineering - who still manage impressive results, in their own ways, and would argue forever with others on how theirs is the best way.

[VFR750R]

Ok, I don't know what 'professional' porters are, cause I have yet to find one that can maximize a ports potential in one shot, or a dozen...or even at all. Even with CFD there is no single solution, no single generalization about ports one can make. There are generalizations like a straighter port flows more, or bigger valves flow more, but in 10/10ths land it really doesn't mean squat. We've learned more about ports in the last year then I think we ever have known, and it's opened the doors to some really cool and exciting stuff but testing takes so long and requires very meticulous testing methods to ensure you get the right result. We measure changes by a single hp. At 800hp that's .125%. We haven't really gotten into surface finish because we're still making wholesale changes to port shape. Slits on the floor of one port might not be needed or might hurt another port.


In real racing, engineering is a method of testing, and a logical means of moving ahead. A lot of information is mythical out there. I'm not going to say smoother is better, I just said I like smoother.

For every guy who tells you he did a test where he smoothed the port walls and it picked up there is a guy who did the same test the other way and it was better. Most of the time its repeatibility issues and the fact that to make a port smooth or rough one must remove material and we've found moving a wall just .010" to have a measurable effect on power. Almost nobody spends the time or money to do a test multiple times or an A-B-A test. 2 years ago we had hand porters blend in the port seat line; when we had the CNC machine do it our hp variance went from 10hp engine to engine to 2hp.

The difference between a good head porter and a bad one is the method they use to do the developement.

[Wesley]

I didn't mean to step on toes.

I agree that even the best porters cannot "optimize" a port just because of the sheer amount of variables that can be tweaked - approaching it from a different way can still net the same power.

Thats sort of what I was trying to say, but I didn't do a very good job.

[VFR750R]

you didn't step on toes, I'm just a (as they say around here) damn yankee.

I'm just afraid of being taken slightly out of context. Just because I desire smooth ports doesn't mean that every other form of racing wouldn't benefit from rough ports. And then I backed it up with an explanation of how complicated port work is. Which really 'flow' is one last frontiers in Mechanical engineering. The most is being learned about it, because it is so complex.

[John Stimpson]

I don't subscribe to the polished port philosophy at all.

I've spent a lot of time on a flow bench, and a rough (I leave them at 50 grit) port with no mod other than a polish almost never flows more.

What was this talk about a bigger boundary layer caused by roughness? Smoother walls promote the build up of a boundary layer, and its common practice to induce turbulence at the wall to "trip" over the BL.

Food for thought...

Why do golf balls have dimples??

[VFR750R]

Well, I might be wrong but 'why do golf balls have dimples' means nothing to a port. How many 'aero' devices other than a golf ball have dimples? If golf balls have dimples why don't airplane wings have dimples, or race car roofs, or straight pipes of any kind, ect. Dimples are for creating turbulence where there normally wouldn't be; and in the case of a round ball of 1.25 inches in diameter traveling at idk 200mph reduces drag. ON an airplane wing you WANT the flow to stay attached. Why or why wouldn't you want flow to stay attached around a short turn? I don't know. Again it's a case by case basis. On a motorcycle port with no distinct short turn, probably no difference either way, but ports with forced geometry may like it one way or another.

Food for thought...

There's cases of supersonic wings where there is a distinct step 2/3rds or so down the top of the wing after the largest cross section where the flow will not stay attached. The step (of specific size) is designed to reenergize the flow and get the flow to reattach later merging with air from the bottom of the wing. It's a case where at subsonic speeds its a bad design, the step is unneeded and causes the flow to seperate at the step, decreasing lift and increasing drag. But at supersonic speeds the flow is going to seperate anyways and so it is the best design for that case.

[VFR750R]

Just a correction as I went back to check what I said, it's not supersonic wings specifically. It's relatively low Reynolds number (for a wing) high L/D.


picture
http://www3.imperial.ac.uk/portal/pls/portallive/docs/1/6335906.PNG

I also found some stuff dealing with the same phenomena using turbulators and vortex generators to achieve similar results.

[Mike Flitcraft]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by KU_Racing:
My stance on this topic is the smoother, the better. Im sure other people will agree or disagree. </div></BLOCKQUOTE>

In higher RPM's I'd agree (take a nod back to Smokey Yunick and some of his tricks to get around the rules of NASCAR a ways back). However, fuel puddling is always a concern of mine.

A mirror finish I don't believe is a very good idea to do, but something finished with say, 120ish grit, would probably be pretty dang good.

[KU_Racing]

I tend to share your fuel puddling worries, but in a road racing application (which is how i think of a formula car, even though its probably wrong) where the intake is at a high temperature and the engine is turning very high rpm most of the time, I would readily sacrifice the low end flow and puddle concerns to make power up top. Like I said, and like almost everyone on this topic has said, you can get as many answers to this question as there are head porters or intake designers.

[Wesley]

I would say that if you left the as-cast (or your grit of choice) finish in the cylinder head runners you would take care of pooling problems, as fuel doesn't spend a whole lot of time in the runners of the intake manifold, at least on the cars that I've seen.

Once you start getting fuel pooling on the backside of the valve as you exceed the injectors pulsewidth, a little turbulence sure wouldn't hurt. Though I would think the sudden localized drop in pressure as the valve opens would take care of mixing up the pooled fuel.

[markocosic]

The consensus (or lack of) is about what I figured! http://fsae.com/groupee_common/emoticons/icon_smile.gif

As the things are dead straight right up to the head and converge from start to the intake valves, and from the start of the head to the valve itself is as-cast, the surface finish would appear fairly irrelevant in the grand scheme of things.

Mismatch in both 'runner-old throttle body remnants' diameters and 'runner-air horn' diameters has been taken out with a 120 grit flapwheel (80 grit or ortary burrs on alloy - far too aggressive!) and left as that.

For what it's worth, budget on 30 minutes per runner with a flapwheel if using the spun Jenvey intake trumpets/air horns/bellmouths - they aren't quite as dimensionally accurate or round as you might hope. (all undersize though - so presumably designed for finishing to fit)

[Bete Noir]

If a carburettor is used, the intake manifold walls can stand a rougher or as-cast finish to prevent fuel acculation and fuel rivlets. If port injection is used, the finish can be very smooth to improve air flow.
If the intake runner is straight, the surface finish is of little consequence since boundary layer effects predict very little fluid motion along the wall surface.
If the intake runner is curved, the surface finish should be smooth as possible to reduce head losses.
The issues of pipes and surface roughness has been very well covered in most hydraulic flow texts, for about 75 years.
Remember that pressure waves seek the shortest path while mass flow particles seek the straightest line.