I'm a software guy, so I'm a bit out of my depth here. It's been very interesting to peruse this forum and see the complexity of the issues involved in pneumatic conveyance systems.
Background: My company is building a manufacturing process which will take bulk powders from super-sacks, through one common reservoir, through pipes with diverter valves into multiple reservoirs on different manufacturing lines. I think the term 'dilute phase pneumatic conveyance' applies here. From there, the powders are dropped by filling systems into the product, which is a canister containing layers of different powders, similar to a water filter.
Problem: The application is medical. We would like to record in our trace system, as accurately as we can, what lot numbers of raw materials are in each manufactured canister. When a new super-sack is loaded into the system, the operator will enter the new lot number (eg: "B"). Our problem then is that the material immediately mixes with the prior lot of the same material (eg: "A"), in the various reservoirs (or 'buffers') and in the piping system. There is no 'line flush' between lots of powders. We can afford to 'over trace' -- meaning that we can record that both lots "A" and "B" are in the product, if there is ANY CHANCE that there are vestiges of lot "A" remaining in the system. But we want to avoid 'under tracing' -- having powder from lot "A" in a canister and not knowing it. Obviously, we can't be 100% certain about this, but we can calculate the 'run out' for lot "A" so as to achieve some level of certainty (say 99.9%). We want to build software that will give us the right lot numbers to put into the "as built configuration" record of each filter canister.
One complication: A single super-sack unloader will feed pipes going to multiple manufacturing lines that run at DIFFERENT RATES. There is a common buffer in front of the diverter valves and a dedicated buffer at each filler machine (each on a different line, running at a different rate). Some lines may be down while others are operating. Some of the reservoirs may have a stirring apparatus. We have detailed knowledge (through the control system) of the weight of powder dispensed by each filler machine, as a function of time.
So finally the question: Is this a problem that has been studied? Is there a reference that I could use that would explain how to approach it?
Thanks in advance for any help you can provide,
Barry
Illustration added by Administrator as an example only:
Tutorial? on Lot Traceability in a Pneumatic Conveyance System
Hi folks,
I'm a software guy, so I'm a bit out of my depth here. It's been very interesting to peruse this forum and see the complexity of the issues involved in pneumatic conveyance systems.
Background: My company is building a manufacturing process which will take bulk powders from super-sacks, through one common reservoir, through pipes with diverter valves into multiple reservoirs on different manufacturing lines. I think the term 'dilute phase pneumatic conveyance' applies here. From there, the powders are dropped by filling systems into the product, which is a canister containing layers of different powders, similar to a water filter.
Problem: The application is medical. We would like to record in our trace system, as accurately as we can, what lot numbers of raw materials are in each manufactured canister. When a new super-sack is loaded into the system, the operator will enter the new lot number (eg: "B"). Our problem then is that the material immediately mixes with the prior lot of the same material (eg: "A"), in the various reservoirs (or 'buffers') and in the piping system. There is no 'line flush' between lots of powders. We can afford to 'over trace' -- meaning that we can record that both lots "A" and "B" are in the product, if there is ANY CHANCE that there are vestiges of lot "A" remaining in the system. But we want to avoid 'under tracing' -- having powder from lot "A" in a canister and not knowing it. Obviously, we can't be 100% certain about this, but we can calculate the 'run out' for lot "A" so as to achieve some level of certainty (say 99.9%). We want to build software that will give us the right lot numbers to put into the "as built configuration" record of each filter canister.
One complication: A single super-sack unloader will feed pipes going to multiple manufacturing lines that run at DIFFERENT RATES. There is a common buffer in front of the diverter valves and a dedicated buffer at each filler machine (each on a different line, running at a different rate). Some lines may be down while others are operating. Some of the reservoirs may have a stirring apparatus. We have detailed knowledge (through the control system) of the weight of powder dispensed by each filler machine, as a function of time.
So finally the question: Is this a problem that has been studied? Is there a reference that I could use that would explain how to approach it?
Thanks in advance for any help you can provide,
Barry
Illustration added by Administrator as an example only:
Traceability ■