Measuring Injection Molding Quality using Weight of final assembly

[pabloquintana]

We have an injection molding shop that produce the plastic parts we require to encapsulate our electronic boards.

I can't remember where but I read somewhere there is a common quality control technique to weight the final product. A lot of defects can be traced to weight.

That could be used to monitor with SPC charts of variables to react faster.

Anyone with experience in this field.

Thanks. Pablo.

 

[Howard Atkins]

Injection moulding is a problem for SPC.
One problem is that a lot of the materials have a shrinkage over time and this affects the measurement of the parts in relation to the time from injection.
This also means that if you are using SPC to get an immediate result you are in fact going to measure an incorrect dimension.
I have argued this successfully with many customers that wanted SPC.
Weight is a good parameter to use as it gives an immediate picture. More than that the nature of non conformances in plastic injection is that the vast majority are visual type defects, flash, non filling, inserts blocked etc, and here a small change of weight will show items that are sometimes difficult to see by eye. It also shows the functioning of the other parameters as the amount of packing of the resin in the part.
There is though a problem because the density of the plastic changes , within tolerances of the producer, the weight can change between lots of resin. This means that there is no absolute but the target weight must be decided on each injection.
In my view SPC by weight in injection moulding is the best method if you want an on line tool.
It is possible to perform a capability study in dimensions but I would recommend waiting 24 hours for this.

 

[bmccabe - 2006]

I like your thinking here. One caution: I've found that unless the material (or sum of materials there of) has a specific density less than say .25 lbs/in^3, the variability from measurement error will prove unsatisfactory. Also, the total weight is a factor. Small quantities are just plane tough to measure.

Just my opinion, I have no real proof.

 

[pabloquintana]

Injection moulding is a problem for SPC.
...but the target weight must be decided on each injection.

How would you decide which target? Any formula?

How can you relate flash marks, sink marks and other to weight? Is there any rule of thumb? For example, if it has flashes weight increases, if it has sink marks wieight decreases.

Thanks guys.

 

[D.Scott]

How would you decide which target? Any formula?

How can you relate flash marks, sink marks and other to weight? Is there any rule of thumb? For example, if it has flashes weight increases, if it has sink marks wieight decreases.

Thanks guys.

Very true, which is why we don't use weight. We use dimensional studies (vision system) to determine acceptability. If SPC or capability studies are required, we use the data from the vision system.

Dave

 

[bmccabe - 2006]

How would you decide which target? Any formula?

How can you relate flash marks, sink marks and other to weight? Is there any rule of thumb? For example, if it has flashes weight increases, if it has sink marks wieight decreases.

Thanks guys.

Sorry, I can't help you here. I know sink marks are likely caused by air pockets (voids) in the material. The plastic shrinks when it cures, which cause the depressions. Target weight and tolerance would be impossible for me to speculate. And you'll likely have long sold the job before any meaningful test could be completed. Flash and voids will likely average to an insignificant quantity. The weight of the components, and coper in the PCB are the higher density items, and probably 90% of the weight. Shoot from the hip?

 

[Howard Atkins]

How would you decide which target? Any formula?

How can you relate flash marks, sink marks and other to weight? Is there any rule of thumb? For example, if it has flashes weight increases, if it has sink marks wieight decreases.

Thanks guys.

This is the reason why the system works, I have done experements which showed very small changes in weight , 0.03g and there was flash.
This is why I think that the weight is a good parameter. It is worth while trying to do a pilot to see how it works in your situation.
Dave, not every one has a vision system, and does it screen all sides?
As to target this is part of the setup procedure, the basic weight is know and you know the difference in density so this is easy to set the base line. The weight in fact shows the amount of packing, how do you know that with out weight?
I have not done this for some 10 years or more. I have a paper that I wrote on this I will try and find it and post it.
Are you talking about weight after encapsulation,if so then bmccabe is correct the weight instability of the PCB will far out way any change in the plastic

 

[Dave Dunn]

Most visual aspects are not going to be reliably caught by weight. Minor flash where tolerances are critical (.1mm in many of our parts) is a negligible amount. Sink generally will be affected by poor cooling as well as low shot volume. Contamination, gas burn, warp, etc. are all visual aspects that will not affect the weight of the part significantly.

Weight can be a good tool to catch more severe problems: short shots, broken tooling, etc. but weight alone will not tell you if your parts are good, only that there is a change in the part mass. What that change means is what you will have to determine with further examination.

 

[Howard Atkins]

Most visual aspects are not going to be reliably caught by weight. Minor flash where tolerances are critical (.1mm in many of our parts) is a negligible amount. Sink generally will be affected by poor cooling as well as low shot volume. Contamination, gas burn, warp, etc. are all visual aspects that will not affect the weight of the part significantly.

Weight can be a good tool to catch more severe problems: short shots, broken tooling, etc. but weight alone will not tell you if your parts are good, only that there is a change in the part mass. What that change means is what you will have to determine with further examination.

I don't agree
I have actually seen discovery of flash by a change of 0.1g in 10 grams.
It all depends on how stable your process.
The only way to know is to try.
It will certainly light a warning which is the main purpose
Obviously the reason for the weight change needs to be found but that is needed in any system

 

[pabloquintana]

Are you talking about weight after encapsulation,if so then bmccabe is correct the weight instability of the PCB will far out way any change in the plastic

I'm talking about the plastic only, without anything in it. What I'm trying to do is setup a more effective quality control system, plus the prevention which is monitoring the machine parameters.

My inspectors will take long time trying to decide if this flash or that splash will be OK or not. Too much subjectivity in. :truce:

Any other ideas?

Pablo