PDA

View Full Version : Danfoss Scroll Weld Failure, SZ380A4CBB.







ryanmj9
19-09-2017, 02:12 AM
Hi All,

I have been experiencing a string of very abnormal compressor failures. Photos attached.

To date two out of eight compressors have failed with a life of 4 years, another is showing a noise resembling weld partially broken creating a vibration/rattle.
The chillers are operating perfectly confirming no flood back or oil issues. Compressors are mounted on correct rubber mounts. However the internal scroll bracket welds appear to be failing, one compressor at a time needing to be replaced.

Has anyone seen this issue before? If so any ideas of cause?
I believe the issue is a poor quality weld from the manufacturing process, as the compressor itself is spotless for faults.

1498014981149821498314984

Grizzly
19-09-2017, 05:36 AM
Hi Ryan
Welcome to the forum.
A very interesting fault there!
I wonder if anyone else has had a similar issue?
When running are these comps showing any signs of resonance emitting from them?

Over to our resident Danfoss expert I think.

Keep us posted please.
Grizzly

Glenn Moore
19-09-2017, 01:49 PM
Hi Ryan
Your photos are not very clear unfortunately so not sure what model and where these were manufactured but certainly welds should not fracture as shown unless there is a production error of some kind.
I recently cut open 3 Scrolls from a site in London all failed just out of warranty. When I inspected them 2 of them had the motor lead wrapped around the oil pump tube and had shorted due to the insulation being worn through by the spinning pump tube. The lead had been incorrectly routed through the motor shield onto the internal plug. These were obvious manufacturing errors and I advised the customer to pursue the supplier for recompense via the manufacturer.
On another site they had lost 6 compressors again this was due to a manufacturing error were the lower bearing bolts had come loose causing the rotor to clash with the stator. This was a known fault by the manufacturer who had already rectified the fault, but again a manufacturing error
I personally would get your local supplier to contact the manufacturers direct with a site report and good quality pictures and demand a failure report from them as this could be an epidemic fault issue which the production team should be made aware of. Often to many manufacturing errors are covered up due to single site issues but when the site has several similar failures the "bells" should start ringing. It may be the crankshaft counterweight has come loose and moved causing vibration which has contributed to the failure, but I would also go to your local University or Technical college to see if they have a welding specialist who could analyse the weld condition to see if there is a weld fault or a vibration or such, site fault.
Danfoss have a huge quality resource to investigate component failures and recommend you get them involved as you may end up with all compressors failing . Note all compressor serial no's to see if they are of similar date of manufacture as there may have been a fault with the welding machine on the production line, but make sure the system functions correctly as well.

Even though they are out of warranty they should not fail due to weak or poor quality welding.
A few years ago Maneurop started reinforcing the weld on the feet of the MTZ80 machines due to the force of the motor stop/ starting on close control systems , breaking the mount feet off causing loss of oil and gas . The mod solved the problem even though the issue really was due to excessive starts so the factory do listen to problem issues when they are informed about them.
No company intentionally manufactures a bad component but sometimes errors do occur but all to often failures are simply replaced and not reported back ,so the factory carries on thinking they are making a good product because they are totally unaware that they are failing in the field.
Please send better quality pics and compressor details and I'll investigate the issue from this end to see if there are any known issues with the model involved

chemi-cool
19-09-2017, 03:04 PM
Hi Glenn,
Interesting incident, all the details you are asking for are on the label picture.

ryanmj9
19-09-2017, 11:05 PM
Thanks Glenn,
Unfortunately the compressors have now been discarded of with the engineering company who assessed the failure cause.
In my opinion the photos I have uploaded clearly show that the welds had not correctly penetrated the outer shell of the compressor. However I have been in contact with Danfoss Australia, they directed blame towards Liquid Flood-back although for the last 3 years I am positive the systems have never seen these conditions. Not to mention the report from assessment engineer shows the bearing surfaces to be immaculate.

Under general operation the compressors are smooth and very normal vibration. As soon as the weld lets go the compressor turns into a bucking bull, this intense vibration has caused pipework to fail releasing 70kg of refrigerant to atmosphere on both compressor failures. It is a worrying issue, not knowing how many compressors are going to fail, no symptoms until its to late and no preventative measures available.

14987

Glenn Moore
20-09-2017, 07:28 PM
Hi Ryan
Have made contact with my old contacts at Danfoss who has requested details of the system etc and any more pics you may have. I have now had a good look at your pics and I would be very concerned at the state of the weld fractures. If I hear anything I'll get back to you, but if you can give me any more details that I can pass on that would certainly help especially the cycling rate and the system set up
Glenn

ryanmj9
20-09-2017, 10:42 PM
Much appreciated Glenn.

We have two Ciat (LGN 1200Z R407c) water chillers, with an arrangement of 2 circuits per chiller made up of two compressors per circuit.
Both chiller units have had the compressor failures.

In regards to the cycle rates, there is a constant load requiring a 6c leaving water temperature. On average run time per compressor is 20mins on - 20mins off. Only one chiller runs at a time, with 100% redundancy from other chiller. Normally only using 50% capacity of a chiller at any given time, with load shedding it allows compressors to cycle off for approximately 1.5hours after a 20 minute on period.

cheers,