Rifling Of Manifolds, Dumps And Exhaust

[Johnny]

Hey guys,

dunno if this idea has been tried or not but im my years of firearms and shooting, rifling of the barrel is the most important technological step in ensuring the highest gas velocity for the projectile.

Pre-rifling, the gun barrels were smooth; the projectile was low velocity and more subjected to wind resistance which affected velocity.

Anyways, has anyone tried rifling each exhaust runner in the manifold to turbo? In my eyes this would increase gas velocity and prevent turbulance, thus increasing spool

Currently, we use a split dump to prevent turbulance when the gate opens.

To put some images in your head, why do american footballers throw the ball in a spiral? to stabilize, increase the speed and distance of the ball - try throwing the ball without spinning on a windy day: it wont go far, fast and will drift in the wind.

I got this idea when reading about the venturi made by that John S guy in the r34 gtr build.

Soooooo anyways why not rifle all the exhaust runners, the dump pipe and wastegate pipe

Thanks for reading guys, just would like to hear some feedback and throw ann idea out there.

-Johnny

[R31Nismoid]

The two things you forget in your examples.

They are both objects trying to MOVE through air.

Not air moving in a single direction due to pressure.

I'm sure if there was any type of gain - it would be a 'new must have' product. Even if it didnt do anything people would still probably buy it blindly if the marketing was good enough

[karma_syke]

this reminds me of that 'vortex' thing that was advertised on late night TV only it was on the intake side. Basically it was a cone with some fins that supposedly twisted the air, i think later it was debunked on some show can't remember though.

[gotRICE?]

Iv always wondered if you welded a little piece of steel into a stock 20/25 manifold where each set of 3 runners meet you could have a basic twin scroll setup for sweet FA money.. anyone tried that?

Edited August 6, 2011 by gotRICE?

[GTSBoy]

Rifling was not to improve gas velocity. Rifling was to spin the bullet so as to stabilise it in flight. Totally different.

As such, the idea is not applicable to trying to improve gas flow inside a pipe.

Worse, spinning the gas will, in fact, increase drag against the wall, because you will increase the gas velocity against the wall, even if you do not change the superficial velocity. This is dead easy to understand. It is exactly the same concept as the three sides of a right angled triangle. If the long side of the right angled triangle represents the velocity along the axial direction of the pipe, and the shorter side of the triangle represents the tangential velocity (the spin component) then the hypotenuse represents the vector sum of these two velocities, and it is of course greater than either of them. This is the velocity that is working against the pipe wall. This is the velocity that causes surface friction. More drag is the result, which then leads to the logical follow ons of increased pressure drop and therefore decreased flow potential.

[Johnny]

Interesting replies, like said its just an idea that popped into my head; im no engineer.

please keep the elaborations coming!

There was a guy on here who had an idea of using boost pressure to pressurise a tank to spray a methanol/water mist into the motor - similar to the electric versions; people like him i admire very much as they like to think of new ideas which may not work but may have some elements which can be utilised.

I would love to see someone measure the flow compared to the tried and tested smooth pipes just to see.

Can see im a curious person ey? lolz

-Johnny

[GTSBoy]

I would love to see someone measure the flow compared to the tried and tested smooth pipes just to see.

Um, sure. Just go ask any 1st year engineering student (well, mech, or chem, not those useless elec or civil losers) to show you the prac results of one of their flow in conduits pracs. It's all been done before. By 17 year olds.

By the same token, using boost pressure to pressurise the water tank of an injection system goes back to the 50s, if not earlier.

cheers

[The Mafia]

Rifling would be a waste if time and great expense, because you aren't realizing that it's for the projectile, not the gas.

Also, using boost pressure for water methanol is another waste of time. You need the wmi at low levels of boost when the car is under the most load. Without much pressure, you aren't going to atomise the the mix anywhere near enough, so it's not goin to work.

My system worked well at 150 psi, the mix was like steam. Anything under 50psi and it dribbled, like the shit your posting.

Common sense, there is a lack of it these days. People need to realize these ideas would be utilised by BMW, Ferrari, Nissan, etc if they were so great. Why? Because they have millions for R&D, and much smarter people than a few kids on a forum.

[T04GTR]

good idea. but not practical. you forget the boundry layer, also you do not want the air to be twisting as you will loose velocity, also you will probably remove the hemholtz resonance effect aswell. turbulance is wanted only just before the valve for mixing the fuel as it is injected onto the closed intake valve. turbulance is also wanted in the plenum to alow a more evan distribution. but the runner needs to be a straight shot for the pressure waves to bounce up and down thru it.

[wolverine]

I believe old school BMW F1 cars used some fine ridges on the internal radius of intercooler piping. Whether this somehow enhanced a boundary layer effect or slightly disrupted flow to even velocity around bends.... beats me.

[GTSBoy]

The smart thing to do there would have been simply to put turning vanes inside the elbows. Odd surface treatments can help to force a flow to detach (in a reliable location, instead of wherever and whenever it feels like) or to help to keep it attached in an area where it might be prone to detachment, or to help it reattach at some later point. Any and all of these things can help to stop energy going to waste (increased pressure drop being the main result of wasted energy in flowing flowing fluids)....but the effects are quite small. The best way to keep flow evenly distributed as it goes around a bend is simply to install turning vanes. This is a little bit difficult to do (shaping and profiling them so that they sit properly inside a round cross section elbow), but that's nothing for an F1 team. They could emply a whole village of Indian tinsmiths to do the cutting out for them.

[Johnny]

The two things you forget in your examples.

They are both objects trying to MOVE through air.

Not air moving in a single direction due to pressure.

I'm sure if there was any type of gain - it would be a 'new must have' product. Even if it didnt do anything people would still probably buy it blindly if the marketing was good enough

lolz yeah i can picture the ebay advert now "spool your turbo 5000rpm earlier"

when you say both object moving through air, not in single direction - what do u mean ash? I dont understand that part?

[D_Stirls]

I've seen the tops on inlet valves grooved/ridged with a spiral pattern but that had more to do with getting better distribution in the cylinder than increasing air flow and i have only seen it on 2 valve engines.

[GTSBoy]

when you say both object moving through air, not in single direction - what do u mean ash? I dont understand that part?

He's talking about the bullet and the football being solid objects moving through a fluid, whereas the gas flows in exhaust pipes etc that you're talkign about are fluid flows INSIDE conduits.

In your object through fluid examples, the only reason to spin the object around its longitudinal axis is to give it spin stabilisation. Spinning it evens out the areodynamic effects of the unevenness of the shape (even bullets aren't perfectly round, footballs have laces) and the gyroscopic action also helps to keep the pointy end facing directly into the flow. The spin iteslf does not directly reduce drag on the object (well, not skin drag) but it does reduce form drag because of the abovementioned improvment of alignment of the spin axis of the object with the direction of travel. That's an indirect but desirable effect of spinning it.

But now, move inside a pipe and impart a tangential velocity component onto a gas. What exactly do you think is going to happen? All that happens is that for the same forward velocity (ie total flow rate) you have now added a tangential velocity component, increasing the total velocity against the wall and so increasing skin drag, exactly as per my first post. There is no benefit to be ganied from this.

[T04GTR]

with the early f1 intercooler comment. that is there to break up the flow to evanly distribute the charge over the whole core area rather than blowing on one portion of it. once the air is in the runner it needs to be as laminar as posible. because the air is not flowing in 1 direction. it starts and stops with every engine cycle.

[rob82]

Iv always wondered if you welded a little piece of steel into a stock 20/25 manifold where each set of 3 runners meet you could have a basic twin scroll setup for sweet FA money.. anyone tried that?

The factory mps exhaust manifold has a swing valve at the exit of it's manifold.

[MrStabby]

with the early f1 intercooler comment. that is there to break up the flow to evanly distribute the charge over the whole core area rather than blowing on one portion of it. once the air is in the runner it needs to be as laminar as posible. because the air is not flowing in 1 direction. it starts and stops with every engine cycle.

But what about boundary layer separation? Rough surfaces induce a little turbulence which helps the boundary layer stay "connected" with the surface, which I assume would assist in getting flow into the top and bottom runners of an intercooler. cf golf ball dimples.

Also, I can't see why air would start and stop with every engine cycle - laminar or turbulent. That doesn't sound right.

[R32 TT]

The two things you forget in your examples.

They are both objects trying to MOVE through air.

Not air moving in a single direction due to pressure.

haha! But grasshopper.. is the gas moving through the exhaust, or is the gas standing still and exhaust moving over the gas...?

[GTSBoy]

The air doesn't stop and start. It is compressible, so it keeps moving forwards in upstream pipework but compresses by conversion of forward momentum to static pressure as the inlet valves close. That propagates a pressure pulse upstream at the local sonic velocity. The flow is pulsatory, and therefor not completely continuous, but neither does it actually stop (except in the region immediately upstream of the inlet valve I suppose). Once you go further upstream of any individual inlet port, you are now into a region where the pulsing flow is feeding into mulitple cylinders - meaning that when one valve is shut, there's another one open or opening soon for the flow to keep heading towards. So intercooler piping is the best case part of the inlet tract for being the most continuous. This all makes it rather difficult to do any sort of analysis on actual flows in these sorts of systems.

Also, you guys need to understand the difference between the terms laminar, turbulent and well developed flow.

There is a very strict definition of laminar. Quite literally it means that if you calculate the Reynold's number for the flow (which is equal to D.v.rho/mu - where D is the local diameter of interest in metres, v is the velocity in m/s, rho is the fluid density in kg/m3 and mu is the viscosity of the fluid in Pa.s (Pascal seconds)) and it comes out to be less than 2000, then the flow will be laminar. Laminr means that the viscous forces are dominating the flow (over the inertial forces). What that means is that the flow lines are all parallel, there is no mixing across the cross section of the fluid flowing in the duct. Thick oil flowing slowly in a 1" tube is a good visual example of laminar flow. If you were to watch that flow, and put some tracer beads in it, you would see all the beads following nice straight paths, parallel to the walls of the tube. The ones near the centre would be going the fastest and there would be a velocity profile from the centre out to the edges, where at each increase in distance out from the centre the velocity would be slightly less, until at the wall the fluid is completely stationary.

If you put a disturbance into a laminar flow, say a step change in pipe size or a vortex generator (any lump protruding into the flow) you may create a local transition to turbulence for a very small distance downstream of the disturbance, however laminar flow will reassert itself fairly quickly. If you have a bend in a pipe with a fluid flowing laminarly inside it, the fluid will generally go around the bend very smoothly and with good behaviour. That is unless the Reynold's number is high enough that the disturbance caused by the elbow is enough to cause the flow to become turbulent as it goes through the bend - after which it will of course go back to being laminar fairly quickly.

There is also a very strict definition of turbulent. A flow is generally turbulent if the Reynold's number is over 4000. Between 2000 and 4000 is generally called the transitional flow region and it is possible to have either laminar flow or turbulent flow, depending on the geometry and fluid properties. It doesn't take much to make a fluid that is flowing laminarly in the transitional Reynold's number region to trip up into turbulent flow though. And it is also possible under fairly specific circumstances to maintain laminar flow up to somewhat higher Reynold's numbers. However, this is not important, so forget it. I only put it in in case someone buts in with that point (in a misinformed fashion) at some later point.

What is important about turbulent flow is that it does not describe how well behaved the flow is, in terms of separations or crowding the outside of bends or whatever. What it really describes is that the flow now has enough energy in it for the inertial forces to overcome the viscous forces. Now the fluid's viscosity is not enough to hold the flow together. So what happens is that the flowline of any individual molecule of the flouid will not be nice and straight and parallel to the flow direction. Each individual unit of fluid will have, at any given moment, a completely random velocity. It might be forward, backward or sideways relative to the bulk flow. It might be a small fraction of the average velocity of the bulk flow or it might be many times the average velocity. The sum of all the velocities of all the particles will add up to the average velocity. Now, all these random velocities are part of a set of eddies that are swirling around in the flow. The higher the Reynold's number, the more energy is crammed into the fluid and so the smaller the scale of the smallest eddies. It is quite fractal. Once a flow is turbulent you can do things to visualise the eddies, and you can see the largest eddies are a certain size. If you make the flow faster (so as to increase the Reynold's number) then the largest eddies appear to be the same size, but now you have smaller eddies hiding inside the larger ones. More speed, more energy, more smaller eddies at continuingly smaller scales hiding, nested inside each other. I suppose the limit is reached when you get down to the molecular scale.

Now, here's the important part. Pretty much any flow in any part of the inlet tract of a car engine is at sufficiently high Reynold's number to be turbulent. And not just >4000. More like tens or hundreds of thousands. So very very turbulent that it can never, ever be called laminar. Not under any circumstances.

But, it can be called well developed. A turbulent flow, if left to flow in a long enough piece of pipe (say, minimum 10 diameters long downstream of the previous disturbance, like an elbow, but 20 diamaters long is better) will settle down so that the velocity profile is pretty even across the pipe. Turbulent flow does not have the smoothly changing velocity profile across the pipe like laminar flow does. It is pretty much like a flatten bullet head. Essentially a flat velocity profile across the full diameter, but rapidly transitioning from the bulk velocity to the wall velocity (essentially zero) in a very thin region close to the wall. The boundary layer. More on that later. This is well developed flow. But if you have a disturbance like a step change in size, or a prrotrusion in the wall or a bend, then the velocity profile will all go to hell. Compressible fluids will tend to flow around the outside of bends, leaving a low velocity region at the inner radius. Depending on the geometry and fluid properties, it may well even separate from the inner radius. The consequence of that is that there is acceleration and deceleration of the fluid as it enters and leaves the bend (which consumes energy), and the recovery process to regain well developed flow consumes energy. Lost energy equals pressure drop. Almost no part of an inlet tract is suitable for creating well developed flow. The flow is always at some level of being poorly developed. Our attempts to make transitions and inlets and all that as smooth as possible are only ever making things less worse, not keeping it perfect.

Boundary layers. Probably the most complicated part of the whole thing. I'm not going to talk about them much. A rough wall can hold a thicker boundary layer. Contrary to popular myth, this will in fact lead to increased friction losses. The friction factor takes into consideration wll roughness, and more roughness almost always equals more friction and hence pressure drop. But, a thicker boundary layer can hold the flow attached to the wall in places where a thinner boundary layer will cause it to separate. So there are places where a rougher wall will in fact cause less pressure drop. But that's only because it stops the flow breaking out into huge ugly eddies that consumer energy to create and restore.

The golf ball example is a classic myth. The dimples do not reduce drag. A dimpled golf ball will fly further than a non dimpled ball only because the dimples grab a thicker boundary layer, which, coupled with the fact that the golf ball is spinning, causes a greater amount of aerodnamic lift to be generated. The extra lift keeps the ball up for longer, so that it can use more of the initial forward velocity to travel further. If you shoot a non dimpled and a dimpled golf ball at the same velocity and don't spin either of the them, the dimpled ball should fall closer because it has greater skin drag.

[GTSBoy]

haha! But grasshopper.. is the gas moving through the exhaust, or is the gas standing still and exhaust moving over the gas...?

Is that a serious question? Because it doesn't matter if it is or not - the situation is the same regardless. The gas is moving in a space constrained by a surface, instead of outside of a surface.