Can somebody help me with this scavenging dilemma? 4

[Gary_321]

I've made a short 48 second video to show the problem.

Youtube Link

I can see how scavenging would work in a system with very little back pressure, but how does it work in a system with a catalytic converter and mufflers?

 

[MacGyverS2000]

Yes, but the cross-sectional area is 7% less.

I can make a square pipe with a cross-sectional area twice that of a round pipe... you're making the incorrect assumption that I was talking about a square pipe that would fit within a round pipe. That's a bad assumption. A 2" diameter round pipe has a cross-sectional area of pi... but I'm sure you could come up with a rectangular pipe that would meet your ground clearance requirements while still having a cross-sectional area of pi.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[Gary_321]

MacGyver - I think we are talking at cross purposes. The crushed pipe that I am referring to is part of the muffler section...

The crushed section is 200mm long. I made the reasonable assumption that the cross-section is oval.

I measured the wall thickness as 1.5mm, the major width as 71mm and the minor width as 46mm. This gives a cross-sectional area of 2299 sq mm.

The pre-crushed pipe is 59mm OD. It's cross-sectional area is 2463 sq mm.

My modification would be to cut out the crushed section and replace it with a 200mm section of 56mm ID pipe.

RodRico & jgKRI - Here is the data that I have been using for my calculations...

 

[Gary_321]

jgKRI - One of my headers (#6) has a 55mm crushed section that reduces its cross-section from 962 to 800. There's not much that I can do about this, unless I want to convert the car to left hand drive .




Re the bucket test, I found that a triangle created a vortex most easily, but if I suppressed the vortex, it flowed the quickest (marginally). Every triangle test created a vortex. Circle created a vortex the least number of times.

Just for fun, here's the data from the test...

Orifice Shape (mm) - Time (no vortex) - Time (vortex)
Circle (15.6) - 36 - 42
Square (13.8) - 36 - 52
Triangle (18.2) - 33 - 52
Oval (25.6 x 8) - 38 - 48
Oval (37.6 x 5.25) - 34 - 50

 

[MacGyverS2000]

Yes, and my point was why replace the crushed section with more round... if you feel you need more cross-sectional area, just replace it with a rectangular section with the desired area (46mm x "as big as you want"). I think we've established it's not really going to gain you much, but my point still stands.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[Gary_321]

That's an option - and a good one, as it will maintain the ground clearance .

 

[RodRico]

Gary,

I had a moment to do a bit of research, and it only reinforces my view that you likely aren't going to get much gain for all your trouble.

The website

http://www.epi-eng.com

is a *wonderful* resource for all things related to engines. They walk through header theory at

http://www.epi-eng.com/piston_engine_technology/exhaust_system_technology.htm

. Note, however, that every application cited is a high RPM race engine that flows a *lot* of air, does not employ a catalytic converter or muffler, and certainly doesn't use Exhaust Gas Recirculation (EGR) to limit emissions of NOx.

The article at

http://www.hotrod.com/articles/header-basics/

does a pretty good job of explaining how pipe length, diameter, and collector design shift the torque peak around. They even go a bit into how one can shift peaks from different tubes to get a small gain over a wider RPM range. The article very specifically brings up the purpose of the headers, scavenging, and mentions that catalytic converters and mufflers will have the expected detrimental result. The article *doesn't* mention that modern cars can *never* attain anything close to ideal scavenging as the intake charge is *intentionally* corrupted with exhaust via EGR to aid in reducing combustion temperature and NOx.

Finally, there are some free calculators at

http://www.speed-wiz.com/calculations/engine/header-1.htm

, and what appears to be some moderately comprehensive software for $75 at

http://www.maxracesoftware.com/PIPE395.htm

. As the epi-eng article notes, however, professionals use much more complex CFD/thermal analysis to design their exhaust systems, so I can't say whether these calculators are really worth much effort.

I have always been highly skeptical of headers on a comparatively stock engine with EGR, catalytic converters, and mufflers; there's a lot of money being made selling headers into street legal cars (read emission compliant with mufflers), so I suspect few want to mention it. If you must proceed, I hope you find the references above to be useful.

Rod

 

[Gary_321]

RodRICO - I've done a lot of research in the last six months and got a copy of one of the "bibles" (Four-Stroke Performance Tuning" by A. Graham Bell). I've seen epi-eng and hotrod. I think that it was hotrod where I read that flow should be measured between "0.050in after open" and "0.050in before close" and not from "just open" to "just close".

I've not seen the speed-wiz calculator, but I have seen many others. I'll look at the page source and reverse engineer the calculations to see what they have done. I always like to go back to first principles so that I can understand the theory.

I have done calculations based on the measurements of my engine and my calculated tuned-RPMs are too high (by a factor of 10). I'll look at the speed-wiz calculator to see if I can get any sensible results before I start another thread about calculating header lengths (if I need to).

Thanks for the steer

 

[Gary_321]

I think the original question has been answered. I originally thought that there was only one type of scavenging (and that was inertia scavenging).

In the original video, inertia scavenging is possible with the header removed, because the back pressure is low and the flowrate is high. The back pressure in the full system is too high for the 15psi blower to generate a high enough flowrate to promote inertia scavenging.

I think that the wave produced when the valve is opened in the 4-second video can be heard using headphones. Unfortunately, this pressure wave would pass too quickly (and be too low a value) to be measured using a home-made manometer.

I was under the wrong impression that inertia scavenging was what determined header length. I think that I understand wave scavenging, but I've still got some reading to do.

Many thanks to all.

[edit] I've downloaded Speedwiz. It cost $30 and looks to be worth the money, but you need to input a lot of data. Luckily, I've been busy with my crank angle gauge, DTI and micrometer!!

 

[MacGyverS2000]

I like this statement on EPI's main page:

NOTE: All our products are ORGANIC, GLUTEN-FREE, CONTAIN NO GMO's, and will not upset anyone's precious FEELINGS or delicate SENSIBILITIES.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[RodRico]

MacGyverS2000,

Did you see the page on a newly discovered element? It's hilarious!

http://www.epi-eng.com/mechanical_engineering_basics/introduction_to_chemistry.htm

Rod

 

[Gary_321]

I've done some calculations and discovered that the pressure in the cylinder drops from around 125 psi to atmospheric in less than 50 milli-seconds at idle, and around 6 milli-seconds at max rpm.

I was planning to pulse some air from a compressor to test the full exhaust system, but there is no way that I could set up a rig that could open and close a valve in tens of milli-seconds.

I would then have to find some very sensitive pressure measuring equipment that could measure the reversion pulse wave in the other header tubes.

It's the sort of experiment I would love to have done for my thesis, but that was over thirty years ago. Unfortunately, I don't have access to that sort of equipment today.

 

[RodRico]

Gary,

How about putting air at exhaust pressure behind a rotary valve? That doesn't seem *too* hard. Alternatively, you could use open source CAD programs and OpenFoam (

https://www.openfoam.com/)

to simulate what's going on. It would take some learning, but at least it would be free.

Rod

 

[MacGyverS2000]

Or you could, you know, just actually use an engine for your testing. A bit noisy and smelly, but sometimes we have to suffer a bit to get to the final result.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[Gary_321]

I don't think that I would be able to source the equipment for a reasonable cost. I could have a go at making a rotary valve and use an electric drill and compressor. I don't think that I would be able to sense the reversion wave, let alone mesure it.

I've been busy since I downloaded Speed-Wiz. It's a fun tool. I recommend it as a tool for many aspects of car design, not just engine. I can't see the source calculations, but the answers seem reasonable.

There are some problems with it. The answers are imperial and some of the conversions to metric are obviously wrong, so I don't have 100% confidence that the inputs aren't similarly affected. All the inputs are based on you having access to a stripped-down engine and rolling road data, so some of the inputs have to be estimated.

It is very good for adjusting inputs to see their effects.

 

[SomptingGuy]

If you have some (significant) spare money, or access to a good library, excellent books for this subject would be the pair written by Winterbone & Pearson:


Design Techniques for Engine Manifolds: Wave Action Methods for IC Engines Hardcover – 28 Jul 1999
by Desmond E. Winterbone (Author),‎ Richard J. Pearson (Author)

Theory of Engine Manifold Design: Wave Action Methods for IC Engines Hardcover – 29 Sep 2000
by Desmond E. Winterbone (Author),‎ Richard J. Pearson (Author),‎ & 1 more


If you want to capture the effects of exhaust design and layout in a simulation, you will need to at least get hold of a 1-D simulation tool. There are some free options. A google search using terms like "1d" "free" "engine" and "simulation" produces many useful leads. Good luck!

Steve

 

[Gary_321]

Theory of Engine Manifold Design: Wave Action Methods for IC Engines

PDF requested from ResearchGate.

Thanks for the steer