Compensating For Material Variability…

I received this e-mail from a medical molder who is encountering complications with maintaining process controls within their validated standards…

Steve
We a medical molder and monitor and record part weight as a control during production runs. On a particular component the spec is 31.165g +/- 0.045g on a 16 cavity mould and was determined from one validation run. Unfortunately, we are continually having to adjust the ‘Holding Pressure’ to stay within the tolerance and sometimes are having to set a value outside of our verified limits. These adjustments are made at the start of each production run and are generally ok for that run or batch of material. The rest of the process parameters are set to validated point conditions and are identical from run to run.

My question is whether weight is a valid method of control with such tight tolerances and what could cause the weight fluctuation if identical point conditions are used for each run?
Note: After a few follow-up questions it was determined that the part is full during first stage, and changes cannot be made to the shot size, transfer position, or fill time.
My Response
The short answer is… yes, weight can be a great way to control tolerances for a molded part.
Unfortunately, the root of your problem is the fact that the part is completely full during first stage, so the process cannot properly compensate for material variation. In this situation, you must fill the mold completely during first stage to meet the validation requirements. Even with sinks, you are actually packing the part during first stage fill. Since most materials will exhibit viscosity variations of approximately 5-10%… the degree of packing that occurs during first stage fill will vary greatly with material variation. These variations can result from machine changes, ambient temperature changes, humidity changes, as well as the obvious variations in the material.
In the long run, it is much more beneficial to establish a process by separating fill from pack. Filling with a short shot during first stage… and packing out the part during second stage will allow the part to fill a little more or less during first stage. completing fill and packing the out during the second stage will ensure velocity is not being used to perform the packing duties.
In your current process… you have a slight tolerance in your process, and I suggest you do your best to slightly increase the amount of material that enters the mold during fill when the weight is low, and slightly reduce that amount when the part is heavy.
Additional Thoughts
In the long term, you should also try to focus on documenting and validating more process outputs rather than process inputs. When you validate all your process inputs, as well as all your process outputs… you eliminate all your opportunities to compensate for material variation.
-Andy
  1. Steve Stubbs left a comment on 2009/11/06 at 7:02 am

    Andy
    Thanks for the response, particularly the additional comments concerning validating more outputs instead of inputs.
    However I have a query concerning using holding pressure to finish filling the cavity. If the part has a thin wall section wouldn’t the holding pressure required to fill the cavity be excessively high leading to moulded in stress and subsequent dimension or mechanical variation?
    Regards
    Steve Stubbs

  2. Andy Routsis left a comment on 2009/11/06 at 7:58 am

    Good question Steve,
    In such a case, I usually aim for a slight short… one that is barely visible. Really, what you are looking for is a buffer from the high speed velocity controlled injection… and packing.
    On your current part, if you have not validated the actual velocity setting, you could also try using a profile with a faster velocity during the first 85-95% of fill, and and then a slower speed at the end to reduce the speed at which fill begins to pack out the part. Varying these is a little tricky, but it may be a way to maintain the desired weight since you have more control over the end of fill.
    -Andy

Leave a Comment