Thermite reactions and how to avoid them.

[Merrill101]

We landed a huge aluminum job in our shop, the first real aluminum production we're going to run. This led to our discussion on how to prevent a thermite reaction when switching back to our typical HR P&O runners. The obvious answer is to clean the scrap trays and empty the dust collector. What other recommendations do you have to reduce the chances of a thermite fire?

Amada 4kw FLCAJ 4020

 

[jjl1210]

We take out the filters from the dust collector and lightly tap them on the floor to help remove some of the partials. Not sure how much it helps be anything that can prevent a fire is a good thing no matter how small.

 

[Chalked]

Do you put water in your dust collector barrel ? Do you cut a lot of rusty materials. If the answer is no. Then I would tell you while anything is possible the odds and likely hood of you getting a thermite reaction are astronomical. If you look into all the factors that have to be present. Ignition source, air mixture,and material component ratio. It is generally not an issue in most laser shops. We operate in the central united States where the humidity is extremely high. Material will flash rust over night. And cut on average 80% carbon steels and about 30% aluminum. In 14 years have never seen a thermite reaction. But once again. Is it possible Yes. Is it likely No

 

[Foostrap]

You should not under-estimate the possibility of having such a reaction. It is rare, but I know for a fact that it can happen. Make sure to empty your scrap bins and dust collector; make sure no accumulation of aluminum dust is left anywhere before switching back to oxygen cutting. Then you should be fine.

 

[krugtech]

I've seen this happen a few times. On one the impeller in the dust collector got hot enough to start throwing blades and they tried like hell to get out of the housing. Clean the dust collector out and make sure the pulse cleaning is working. a spark arrestor is also a good idea.

Chris Krug

http://krugtech.com/

Maximum Up-time, Minimum BS