Had to leave for a little, updated!
Awesome, thanks for the whole riveting saga in one megapost.
For others reading this - Erik's core issue was wastegate flutter. From the video he sent me it did sound a bit like compressor surge, but with a deeper tone and lower frequency of oscillation.
A good way to think about this type of wastegate issue is by visualizing the forces acting on the gate itself (draw a free-body diagram if you're an enginerd). Below is a handy exploded view from Turbosmart.
Forces acting on the valve are:
- On the bottom face, exhaust backpressure (pre-turbine) times the valve face area - acting to open the valve, pushing left in the image above
- On the top side of the valve face is the post-turbine pressure (or atmospheric if dumped to atmosphere) times valve area minus stem area - pushing right
- On the stem, pushing right (acting to close) is the spring preload force when the gate is closed, or spring compression times spring constant k while opening
- On the stem, pushing left and acting to open the valve, is control pressure times diaphragm area
- Finally, depending on boost control configuration you can have pressure acting on the top of the diaphragm, times diaphragm area, pushing right and acting to close the valve
My take on what was happening below....
In Erik's test case, with both ports open to atmosphere the diaphragm wasn't doing anything at all to open or close the gate. Only the spring preload and the forces on the valve face were in play. With the 4psi spring the preload force is very low. As pre-turbine exhaust backpressure builds up in the turbine housing, only a very low pressure is needed to overcome the spring force and open the gate, releasing all the exhaust energy, killing turbo speed and boost pressure. As the gate opens and boost & engine power fall downward, the pre-turbine pressure would also go down, to the point that the wastegate starts to close again, under spring force. As the wastegate valve is closing the flow area is getting smaller between the valve and seat, which speeds up the exhaust gas velocity, which actually lowers the pressure on the valve face even more through Bernoulli's principle, so the gate slams shut. Now that the gate is shut the exhaust gas energy starts to build again, up to the point that the spring pressure is overcome again, and the valve starts to open a second time. And so on, etc. etc...this happens in cycles until Erik gets tired of it and lifts off the throttle.
The video was interesting for sure and got me thinking about wastegate control stuff that I haven't dealt with in ages.
By going to a smaller wastegate valve diameter the pressure forces on the valve is reduced, because F = pressure x area. Same gas pressure but reduce the valve area and the opening force goes down.