Rough running 912UL

Our 700 hr 912 UL has just had  a set of rings installed.

On the flight back to the airfield from the AME when turning onto base and throttling back at around 3800 - 4000 rpm a severe vibration developed. Dropping the rpm to around 3000 rpm reduced the vibration but not stopping it completely.

On the ground all the basics were checked and no obvious faults were found with the exception of the spark plug colour. The attached images show plugs 1 and 3 a velvet black (very rich) and plugs 2 and 4 a normal light chocolate brown.

- The carbs had been rebuilt in April 2020 and and were pneumatically balanced after the new rings were installed.

- Minimal idling between landing and shut down.

- The floats are the new 'R' type, float with the tab pins on the fuel surface, each pair weigh 6.33g, the float arm is parallel to and measures approximately 10.5 mm from the bottom face of the carb base.

- No restriction on the air filter or inlet tract.

Having found nothing else obviously wrong, a few days later we went for a flight. The conditions were 2/8 cloud cover and moderately turbulent.

During the cruise the engine was smooth, all temps, pressures and flows were normal until we hit an updraft in the cloud shadow. When throttle was reduced the engine started running very rough, sounding like it was running very rich and the fuel flow jumped from 10 l/hr to 20 l/hr . Opening the throttle when exiting the updraft cleared the rough running. This cycle occured each time the throttle was closed and then opened.

We returned the airfield for further checks.

- Spark plugs 1 and 3 were velvet black, plugs 2 and 4 light chocolate brown.

- No leaks found on the outside of the float bowl or fuel in the overflow pipe.

- Fuel level normal in the RHS float bowl.

- The piston in the RHS carb moved freely when lifted.

- The scat hose to the inlet of the RHS carb had a stong smell of fuel.

- No debris was found in the RHS carb float bowl.

What would cause the rough (rich?) running when closing the throttle and why would the fuel flow increase?