This is a common question that comes up a lot on this forum. The model is definitely unstable but tracking down how it is unstable is the real problem. I went through a quick stability diagnosis in a duplicate of your study like this:
1. Go into the study 'Properties'
Turn on 'Use soft springs to stabilize model'
Turn off 'Compute free body forces'
Turn off 'Large displacement mode'
2. Suppress all loads and apply only gravity
3. Mesh/Run the study
4. When the solver stops due to large displacements and prompts to put the study into 'Large Displacement mode' declined this offer. The solver will dump what results it has so far
5. Inspect the displacement plot in the results to find the unrestrained body
The results showed one body was not attached to the rest:
Since you have a Global Bonded contact definition in place already you may wan to create specific bonded contact sets to bond the elbow onto the flange ends.