I have heard anything up to 50 parts is an acceptable level of scrap during startup. My experience has been 10 to 15 parts are an acceptable level but others dispute this as a quantity that is too low. Any thoughts?

I think it depends a lot on the complexity of the part. If you are molding washers you can probably get away with 10 to 15 scrap parts. But if the part has any complexity or the mold requires actions to produce the parts even 50 might be to low. In addition it depends on what is required for capability studies. At Dukane we make equipment to weld plastic parts together by a number of different processes. As a rule of thumb we want a minimum of 50 samples to insure that the process is stable and will be satisfactory in production.

The only way to know this for sure, like many part quality related questions, is to do a test. Donald is correct in stating that it may depend on complexity, other factors are tool construction, material selection and how robust your process is just to name a few. The key question is what is the job of your part and when does your process reach an acceptable level of part quality.

Start up scrap can vary widely as you can see from the first two comments, however, the key is to have a qualified process engineer or a very well trained process technician on staff that can determine the optimum process for that mold using scientific molding techniques. Then, the next time you start that mold up with those parameters, you will get your starting point for scrap. From that point, continuous improvement kicks in and you have to begin eliminating the causes for start up scrap and variation based on the return on investment you get for the effort. Obviously, you don't want to spend a fortune to get rid of a a few scrap parts. The cost much match the benefit.

It depends on your level of patience - some people warm up the mold by producing parts. This makes for a lot of scrap initially, others let the equipment bring the mold to temperature. You still have to allow the injection side to stabilize. My usual rule of thumb is 30 parts or 30 minutes for precision parts. Less precise parts don't need as much time.

The last two comments speak volumes. From a pure process perspective, achieving thermal equilibrium in the tool and the injection barrel will take some time. Bill's note of 30 minutes is valid depending on the size of the press. As Teri mentioned, with some investment in testing, a fairly solid starting point can be developed to reduce scrap. However... from what I have seen recently, with as tight as money has gotten the reality is that you may or may not know how the batch of raw material your boss just dropped on you will react. As I have said before, when you deliberately inject chaos into a process, lower your expectations.

I might also add something on FIRST PART APPROVAL.
1. Optimize your process first, then modify the mold to get it into spec.
2. NEVER, send the 'crown jewels'. Send what will be typical for future production.
3. Once you have an optimized process, don't be silly enough to speed it up. Good parts are better than scrap parts made quickly.

Good comments,
Process stability determines dimensional stability of the parts, more importantly whether ii is single or multicavity based on these criteria it takes more time to get acceptable parts, and agree with earlier comments on complexity of the parts takes more time on fine tuning the process window, atleast average 30 minutes required to get good parts irrespective of number of cavitation