Similar to 20612. It's be around for a few years and is in at least 20 sets so should probably be modeled.
I wanted to do this quickly, but after comparing what we have and the external resources I had a chat with Philo first as it feels like a little can of worms...
Philo's 20612 is based on a LDD Mesh from Digital-Bricks
Looking at 20612 and 1941 from the BI, the location of the perpendiculat bar is one LDU lower...
How do we approach this? keep the official 20612 positions? Might be the best