Welcome • User Guide
ToS • Privacy • Canary
Donate • Bugs • License

©2025 Poal.co

In this image we see a normal shockwave in a pipe, the pressure builds at the inlet, and chokes at the throat as the fluid reaches mach 1. We call it choked because the pressure builds up and the pressure terminates in a shockwave, downstream the pressure drops. Along the wall we see as the pipe widens, the pressure drops at the wall, but raises at the center. The pressure at the center is caused by secondary shockwaves by the supersonic flow along the wall reflecting and adding to the drag caused by the primary shockwave. You can see in the first image the expansion fan is weaker and the reflecting compression waves are also weaker leading to a more uniform flow downstream.

The bottom image where the normal shock is strongest has the strongest compression waves from the expansion fans, creating in effect a constriction in a supersonic flow. In compressible/trans/super sonic flows constrictions raise pressure and slow down airflow. It is only after the bulk airflow passes this constriction that the overall pipe pressure decreases. Had the wall been constructed more carefully, one could largely avoid the compression waves while maintaining the expansion fans that further accelerate a choked flow. This would reduce losses substantially.

In the bottom image if you look closely you can see the recirculation bubble.

In this image we see a normal shockwave in a pipe, the pressure builds at the inlet, and chokes at the throat as the fluid reaches mach 1. We call it choked because the pressure builds up and the pressure terminates in a shockwave, downstream the pressure drops. Along the wall we see as the pipe widens, the pressure drops at the wall, but raises at the center. The pressure at the center is caused by secondary shockwaves by the supersonic flow along the wall reflecting and adding to the drag caused by the primary shockwave. You can see in the first image the expansion fan is weaker and the reflecting compression waves are also weaker leading to a more uniform flow downstream. The bottom image where the normal shock is strongest has the strongest compression waves from the expansion fans, creating in effect a constriction in a supersonic flow. In compressible/trans/super sonic flows constrictions raise pressure and slow down airflow. It is only after the bulk airflow passes this constriction that the overall pipe pressure decreases. Had the wall been constructed more carefully, one could largely avoid the compression waves while maintaining the expansion fans that further accelerate a choked flow. This would reduce losses substantially. In the bottom image if you look closely you can see the recirculation bubble.

(post is archived)

You are currently inside a comment thread.
Click here to see all the comments (8).
[–] • 1 pt

Ah. Yes, Ive had that as well. Airflow must be measured in a non transition zone. Exhausts approaching supersonic speed behave a bit differently, and the loss of temperature and volume is offsettable with tapered pipe and insulation. The boundary layers are fun to play with, I came up with a neat way of dealing with them that worked both low and high rpm with gains on both. The prototypes broke all the normal rules, at some point I may build a few more and patent them.

[–] • 0 pt (edited )

Yep if the temps go down it's easier to get sonic choking, hot air is less dense which is the equivalent of increasing cross sectional area. Hotter temps means it takes more to reach mach 1, so keeping the exhaust nice and hot buys you some wiggle room. Once you get sonic choking, the boundary layer shows you the shape you need to make the wall in order to get max efficiency. You just have to run the thing at the flow rate you want, then capture the boundary layer (PID/Schlerien/CFD), then design the walls to mimic the shape of the BL, then you cut losses dramatically. The flow can just follow the wall curvature instead of creating losses to define its own path of least resistance. You basically make it easy for the flow, and it keeps accelerating.

https://share.southbox.de/f.php?h=1RMD9k4o&p=1

You'd think that it's only a coincidence that this cross sectional schematic looks like the plan form of a bird?

[–] • 1 pt

Not all things are well documented yet… there is ways to trick the flow and reduce greatly the choke points increasing the flow and minimizing the energy required to move the gas. Hindle I think I remember was the first to market a full tapered pipe exhaust, which made it velocity constant over its length. Even small scratches or bumps inside a pipe can have large effects on flow as they create or move pressure waves.

As to air under a vehicle, I use to duct the air underneath into a cooling flow that exited in the negative pressure area near the rear window. Solved a few problems neatly. Many modern vehicles use a flat belly pan and a front air dam to reduce any significant underbody turbulence to the point it isn’t a concern.

Deploying a handful of pressure sensors and mapping the pressures under various conditions is usually a good start to see where things stand.

What parts are hanging in the wind causing drag?

[–] • 1 pt

Strakes to create vortices on the inside of the wheels. I discovered a way to improve aero performance by doing something that looks stupid on the surface. Most aero is focused to getting air around the outside of the wheels, using air screens to reduce turbulence. I went to the inside of the wheel instead, by combining wheel alignment, and strakes, I'm directing air so that it outwashes into the inside of the tire along the brake ducts. I learned via experiments that if you toe out a spinning wheel it creates a strong jetting vortex, so I guide airflow along the inside of the wheel, along with suspension aligment so the wheel sees basically airflow at a 15-20 degree angle depending on where the front wheels are pointing. I don't have as much freedom in the back, but doing this took out all the understeer at the front of the car. I had to weld brackets to the chassis to keep the upper control arms from being ripped off the chassis. I did this on 200 tread wear tires, you need to run slicks to get that kind of loading, so I know I'm on to something.