The photographs presented here are taken from "Mechanics of Fluids" by Duncan, Thom and Young (English Language Book Society, London).

Below are pictures of a 10% thick symmetric aerofoil at 2 degrees angle of attack in a wind tunnel. The flow Mach number is gradually increased from just below critical to supersonic. The growth and development of the shock patterns are very clearly seen.

At 0.7M no shock has formed, implying that M_cr is greater than 0.7. In this and subsequent photographs, the wake is visible as a horizontal dark line starting at the trailing edge.

At 0.75M the first, weak, shock appears on the upper surface at about 20% chord.

In these pictures the shock is growing in strength (the shadow is growing longer and darker) and moving rapidly towards the trailing edge. The dark region just behind the shock at M=0.82 indicates turbulence - the boundary layer is on the verge of separation.

There is no shock yet on the lower surface.

At 0.84M a shock first appears on the lower surface. It is positioned almost directly below the shock on the upper surface. On the upper surface, the shock has grown in strength and the turbulent zone behind it is more prominent, but it has hardly moved. A "crow's foot" has formed at the foot of the shock. This has happened because behind the shock, due to the sudden large rise in pressure, the boundary layer has become quite thick and the shock reflects back and forth in the boundary layer.

At 0.88M both shocks have grown further in strength. The shock on the lower surface has moved closer to the trailing edge while that on the upper surface has not moved. Now, clearly, the boundary layer has separated downstream of the shock on the upper surface. This kind of separation is known as shock induced separation. If it sets in while the shock is still well forward, complete and sudden loss of lift can occur. This phenomenon is called a shock stall.

The shock on the lower surface has almost reached the trailing edge. The shock on the upper surface has finally started moving back, apparently very reluctantly. The "crow's foot" and the separated boundary layer are both more prominent.

Both shocks have now reached the trailing edge. The "crow's feet" are clearly seen as the shocks interact with the thick wake.

The free stream is now supersonic. At 1.10M the shock near the leading edge is so far forward that it is not in the frame.

At 1.18M the upstream shock has moved closer to the leading edge and is visible in the frame. It is almost a normal shock.

And now, a diamond aerofoil 12.5% thick at 0 degrees angle of attack.

The shocks at the leading edge are attached and oblique. The expansion fans can be seen as the relatively light coloured fan-shaped regions centred at the convex corners on both surfaces.