Eric,
I guess that it is time for me to put in my two cents worth on my favorite aero book "The Leading Edge" by Goro Tamai. I know it is about cars that do not run over 60mph but the aero info does translate to what we are interested in and he does address interference drag, especially junction flows, appendages, fillet radius' etc. It also addresses most of the things that are important to land speed racing related to aerodynamics.
I 1997 or 98 my son and I were at the 5 mile mark when Don Vesco came through at 439 mph in his turbine car, at that time the air inlet to the engine was a scoop that was mounted on the top of the car above the boundary layer. The day was somewhat humid and we could see shock waves coming off of the back of the air scoop. Does your statement " yes, bumpy cross sectional area distribution makes lots of shocks." explain what we saw?
Rex
I haven't seen "The Leading Edge", if we see each other at Bonneville this year I'd like to get a look at it. Is it still in print?
The speed of sound at Bonneville under the racing conditions of the various meets of the last half century range from ~1100 fps for the 43F during Blue Flame's return run to ~1150 for a rare 100F day. Given this Turbinator's exit speed of 470 mph (691 fps) would translate into M.60 to M.63. This is certainly high enough to see some areas of abrupt curveture develop supersonic flow. This would lead to compression waves on the front end of these areas as the local flow accelerates beyond M1.0 and expansion waves behind as the supersonic flow turns to fill in the downstream pressure recovery area.
Looking at the pictures on the Vesco site, it appears that the increase in area on the front side of the scoop was smooth enough that any compression waves would be spread out and not visible as an atmospheric disturbance. The back side of the scoop is pretty abrupt, so what was likely seen was the expansion wave from the flow turning to fill in the separation area.
The best visible example of this effect in LSR was the famous TSSC photo. Look at the shock that comes from the back of the engine nozzles: There was 39 ft2 of base drag in this area and it caused a large and VERY high drag expansion wave at a completely different visible angle than the compression waves from the other area-increase features (nose, inlets, cockpit) and the trailing expansion shock at the tail.
On a properly designed supersonic vehicle, not compromised by poor aerodynamics, there should be one compression shock spread over the front half of the vehicle and one expansion shock spread over the back half with hard shocks only at the front and rear. This series of compression-expansion-compression waves on the analysis of the current vehicles should have been fixed long before CFD. Some will say that this is impractical on an LSR vehicle. Not true. I solved that problem 3 years ago and we've been working on other issues since.
Got a little off topic here. Good to have a board just for this subject.