True Professional Engineer
This is a blog to share my experience in manufacturing / industrial engineering and management system setup.
Please feel free to follow.

Tuesday, April 17, 2018

Trouble shooting for mechanically jammed Serpentine oven

This is a recent incident in one of the plant with an Auto-bake Serpentine oven. A picture of the Serpentine oven is downloaded from the Auto-bake website as shown below.

The picture above shown only a Serpentine oven. The advantage of having a Serpentine oven is the same oven drive chain can be lengthen and passing through multi compartment with different environmental condition and make into a full production line, except packaging.

As shown in the sketch below, the production line that was mechanically jammed has an Infeed section, an Oven, a Cooling section and an Outfeed section. The width of the line is approximately 1.5 meter. It has two double pitch roller chains running at both side of the line with carriers hanging in between the two roller chains.

Both roller chains are a few hundred meters in length passing through multi loops as shown in the sketch. Thus the whole line is running with two motors (drive 1 & drive 2). There is a synchronizing encoder in between the two drives, sensing the tension condition of the chain and controlling the speed on drive 2 in related to drive 1. This serves to share the whole line load between the two motors.

The incident started with an alarm, "drive 1 mechanically overloaded". The production team reported the issue to engineering team and they started the trouble shooting process. 

The engineering team first noted that there was no output from the variable frequency drive unit of motor 2. Hence they thought it was the faulty variable frequency drive that causes motor 2 not working which lead to mechanical overload alarm of drive 1.  With this finding, they planned to check into the variable frequency drive unit and replace if necessary.

When I was informed on such issue, I stopped them from any work on the variable frequency drive but ask them to check into the position of synchronizing encoder. I explain to them that the variable frequency drive 2 might have no output if the synchronizing encoder indicates that the section of roller chains driven by motor 2 are tight. The check was carried out and proven right. The synchronizing encoder was fully retracted because the sections of rollers chains for "cooling" was tight.

I ask the team to check on any obvious mechanical jam, especially by the ends of all loops because a straight run of chain normally should not jam. The team finally reported thaty there was no obvious jam and they are running out of idea on how to check further.

Then I decided to manually pull stretches of rollers chain in various sections of the production lines, with the intention to feel if there is any difference in tension. The tensioner on the outfeed end for both roller chains are released before the "feeling process". 

The "feel" of both rollers chains (both sides of the production line) in the cooling section are quite even. The same goes to those roller chains in the oven. When it came to the infeed end, the roller chain on the left hand side seems to be tighter as compare to the right hand side. The difference became more obvious on the return path of the roller chain. The difference was very slight and I decide to get the executives, engineers and supervisors to feel the difference too as a matter of On-Job-Training.

Finally when it came to feeling the roller chains on the return path of the outfeed section, a greater difference in tension was found. With this "feel" action, we finally manage to trace toward the jammed sprocket as shown in the sketch.

Removing baking trays from the section and climbed into the suspection area, we then found the little M6 X 10 Counter sunk screw jamming in one of the sprocket.

With such a small piece of screw jamming in one of the sprocket which is not easily accessible, the tracing technique is important or else it would be difficult to find even though the problem is truely minor. Hence I decided to share it in my blog here and hope that readers who had read this will be able to lear the little practice of "feeling the machine" during trouble shooting.

Thursday, April 5, 2018

Chiller Unit tripped with "No Flow" alarm

On passing by the Chilled water system and glycol chiller units in the plant this morning, I saw the supervisor and technician are working there for some problem faced. On checking with them, I was informed that the chiller unit tripped with "Glycol no flow" alarm. It runs back to normal after resetting.

I decided to ask if they understand what possible caused the tripping and I was informed that the level in the glycol tank is low. With doubt in mind, I decided to go with them to check the glycol tank level and noticed that the level was not "too low" as what they explained.

Hence decided to explain to them on how to trace for the possible root cause on such system trip issue. As I always highlighted to the team, a successful man always slice a large problem into smaller pieces and then tacker them one by one. In no time, the problem will be solved. The same apply to engineering trouble shooting.

Hence the first step is to trace the part of the whole system that giving alarm "glycol no flow to chiller unit". The following snip set of the glycol system related is capture and atteched here for reference.

Using the glycol tank as a starting point, we noted that there are three pumps (PP1, PP2, and PP3) taking their supply from Glycol tank and discharge into a common line to send for chilling into two chiller units (Chiller 1 and Chiller 2) and then back to glycol tank. Two of the pumps will be running at any one time with the third pump on standing by. There is another circuit with two pumps (PP4 and PP5) taking suction from glycol, sending through Chiller 3 and then back to the tank. For this circuit, one pump will be running and the other on standing by too.

After check and confirm the level of glycol in glycol tank is normal. I noticed that the running pumps are PP2, PP3 and PP4. The pressure gauges readings were PG2 zero PSI, PG3 80 PSI and PG4 50 PSI.

Hence my first advice to the technician and supervisor is to ensure all presure gauges are working correctly. The pressure gauge that indicating 0 PSI is definitely faulty or blocked. Assuming pressure Gauge PG4 is correct at 50PSI, then pressure gauge PG3 with pressure of 80 PSI may be faulty too, or there may be some blockage on the line. Hence, I advice them to check on any possibility of partial iced up in the heat exchanger of the running chiller unit. The system may have to be shut down for a short while to let the iced glycol melt.

A further action is required to check the glycol bridge. It may happen that the bridge is too low, or else the glycol should not iced up in the heat exchanger.

After ensuring that both the supervisor and technician understand the explanation above, they were then left to continue with their action to rectify the issue accordingly.

"NO FLOW" alarm is activated by the differential pressure switch PS connected across the inlet and outlet of the heat exchanger. Hence an alarm basically indicate possibly blockage of the heat exchanger. 

Trouble shooting for mechanically jammed Serpentine oven

This is a recent incident in one of the plant with an Auto-bake Serpentine oven. A picture of the Serpentine oven is downloaded from the Aut...