Area-51 R2, overheating

That's not close to overheating the CPU at those temperatures. 

What are the blue screen error codes? 

What critical events are logged in the viewer?  Are the shutdowns coinciding with kernal power events? If yes, this could indicate a failing PSU. 

What does reliability monitor say? 

Have you tried running the built in diagnostics? 

@r72019 is correct. temps seem normal.  Test w a different verified working psu, as when gpu is underpowered by a failing psu, system may lock up(freeze) or BSOD. can also test by removing all ram sticks and insert only one back in DIMM1.  test each ram stick individually.  During the test disconnect hdd/ssd from system.

Okay so not overheating then.

The blue screen erro code is  'WHEA_UNCORRECRABLE_ERROR'  doesn't say anything else.

The viewer shows that the source is 'Kenel-Power' and event ID is '41'.

The shutdowns coinciding with kernal power events yes.

Reliability monirot says that it's a 'Hardware error' called 'LiveKernelEvent' code '141' 

My diagnostics doesn't show anythings wrong.

That could be consistent with a failing PSU, an under powered PSU, or an issue with the GPU.  

If the issue is coming up randomly under idle with no load on the gpu, I might tend to suspect the psu.   Is it all stock equipment, or have you upgraded anything?

You could also try updating your gpu drivers from nvidia if not already.

That stock psu is 1500W hard to get replacement. Try install a very simple video card that takes no gpu power connector to replace the stock gpu. something that has TDP in the teens would do. Test whether that makes any difference. With a small video card you can test the system w any wimpy but verified working psu which can be 100-200W or less. This may help you determine whether the stock psu is the actual culprit and needs to be replaced.

https://support.microsoft.com/en-au/help/4028662/windows-how-to-fix-wheauncorrectableerror

1. The viewer shows that the source is 'Kernel-Power' and event ID is '41'. The shutdowns coinciding with kernel power events yes.

2. Reliability monitor says that it's a 'Hardware error' called 'LiveKernelEvent' code '141'

3. My diagnostics doesn't show anythings wrong.

=====================

1. Normal if the machine crashed, turned-off un-cleanly, etc.

2. Likely (hopefully just) video-drivers.

https://www.dell.com/community/Alienware-Desktops/My-Aurora-R6-The-Adventure-Continues/m-p/7492048/highlight/true#M22717

3. ePSA is great, but only tests hardware up to 90%. Also, its video-card checks are minimal/basic. For example, it does not check high-end/gaming use (3D acceleration) , Direct-X sub-system , etc.

Also, all those other Apps shows crashing or "stopped working". All that needs to be fixed. If updating/re-installing doesn't help ... I think I would just uninstall them for now.

Tried updating different drives and managed to get it to stop restarting. Howeer when I launch a game it now just freezes and doesn't respond so I need to power it down.

Thanks for the help so far and now think t's an issue with the graphics card.

Turns out it is still restarting even when doing nothing on the desk top..... 

If psu is not able to power gpu system will freeze or shut down restart. Back to my previous suggestion, test w simple low TDP gpu (one that does not need gpu power connector) and verified working psu.  Other things that could go wrong would be hdd. For that you could disconnect hdd and testing Windows 10 install app (proceed to beginning steps) and let system sit idle there to watch for stability.  Lastly remove all rams and insert only one stick back, testing each stick individually for effect on system stability.