We understand that with plastic parts, weight is the best method of controlling the process capability. Is there a guideline or standard for defining the tolerance?

The most common method is:
Shot to shot weight variation calculated as (three standard deviations) divided by (average weight) must be less than 1% and should be less than 0,5%. In general it is possible to achieve 0,1-0,4% weight variation using scientific molding.

What is the average operating temperature?
Material RTI value is critical for this
What is the peak operating temperature?
HDT is probably the most critical value
Is it wet or extremely dry?
Water Absorption rates can have an impact on mechanicals. Nylons have some of the highest rates.
Is it exposed to chemicals?
Chemical compatibility specifications must be considered

It's a small plastic injected piece that will be submitted to short period of time (2 seconds) of high temperature (95-105ºC approx.) in every cycle; this high temperature is produced by a filament where electric current passes through (like a house light bulb but hotter) for no more than 2 seconds in the process cycle. A complete process cycle last for about 45 seconds where the piece has time to cool off again. This plastic piece has to work for about 45-50 complete process cycles.

I have used ULTEM (SABIC innovative plastics) PEI (polyetherimide) in high thermal service components such as hard drive parking ramps, insert molded side looking LED optics, motion sensor windows, and missile fuse optics in the past with good success. It also has high lubricity and it is easily used in molds designed for both commercial and engineering grade materials with shrinkage of 0.005 to 0.007in/in.. It comes in a variety of application grades with a 94VO rating.

Injection molding of phenolic has been around most of the 40 years I've been in the industry. Although it's not conventional (same as thermoplastic) equipment it is quite similar. It does take a special screw and a jacketed barrel, the basic concept is the same as thermoplastic injection molding and the same basic machine can be used with the proper modifications to the screw and barrel. In my opinion, this is the easy part and the part that is already being done by REHA, the complicated part is compounding the resin into a molding compound. Compounding is not really something that I would suggest a molder take on.

I believe that is a BUMP in the clamp during the injection phase when molding rubber or silicone. The clamp is allowed to breathe (reduce tonnage) to expel the air trapped inside the mold. The clamp can be opened and closed ever so slightly (i.e. BUMPED) during the filling/pack phase of injection. This can also help to reduce molded in stress. The majority of press manufactures that have accurate control over their clamp movement can add this option to their control for a charge.

The compressional motion can be done by injection compression molding machine itself or separate mechanism linked with traditional injection molding machine. The mechanism is normally hydraulic-designed. It pushes mold plates or necessary cores to compress prior-filled melt then a finished part is made. To complete a process cycle, the mechanism is normally also integrated/programmed with injection molding machine so that its actuation/ending timing signal of compressional motion can be sent from/to injection molding machine.

The most critical element is to have a good design first with special considerations to the structural stiffening members. Then use glass fiber reinforced PP, but remember the injection molding of glass filled PP is different from the injection of unfilled PP even if you are molding the same shape. The stiffness and the strength of the part depend on the fiber orientation pattern throughout the molded part. Such pattern is determined by the location and the type of the injection gate. The fiber length distribution which determines the quality of the compounded materials has huge effect on the creep modulus and hence on the long term strength of the par during application. It is not enough to use glass fiber filled PP without specifying the average fiber length. It is also important to consider the effect of the fiber orientation pattern and the fiber length on the warpage and shrinkage of the molded part which is completely different from those of the unfilled PP molded part.

In higher humidity conditions, apart from the increase in water content in the air in the tool void prior to fill, maybe you're getting condensation at times (if the press is held-up between shots), in which case you will be pushing lots of water into the weld-line. Worth noting, also, is that even if you have the same barrel settings and water temperature and flow settings throughout the year, actual conditions that the material encounters (precise melt temperature at point of entry into the injection mold and actual mold surface temperatures) will vary, because of the R.H. and temperature differences. Hire a thermal imaging camera at relevant times and look at the readings on the parts at point of ejection. You'll see differences, I'm sure.

There are possibilities to employ vacuum-assisted venting (by default, this would pull moisture out with the air) in the injection mold void prior to fill. Also, as I've seen with multi-impression PET preform molding, air-conditioning enclosures around machines.

The important consideration should be end product and what your definition of "REGRIND" is. Reprocessing your own internal scrap (if kept clean and uncontaminated with anything else) is a significantly different question basis than using unknown or unclassified "regrind" from outside sources.

Most folks that are quoting probably don't realize the write UP in material "cost" that should be considered when significantly less percentages of regrind are incorporated. If you DON'T use it up and end up selling it for a few cents on the dollar, the WRITE DOWN is actually a part of the COST of the product you originally bought the material for. And most don't have any idea of what the "true" cost to own of an 80:20 mixture is considering handling, size reduction, inventory holding and shrinkage costs etc are over the life of a business project. This should be part of the 'living business plan' for every product life cycle in your factory.

Rounded pellets usually are hot-cut, cylindrical pellets are usually strand-cut so most likely different extruders were used. This (change in extruders) may effect compounding of the materials and thus properties, but your supplier should have done QC testing to know product is the same. Again, this may be where pellet-to-pellet differences could show up in FTIR analysis. Also, most suppliers don't mold sample test specimens that have knitlines that many production parts contain, and they wouldn't be able to catch differences in batches if you are having knitline issues. Poorly compounded materials (not enough mixing) might show up in drop dart impact test comparisons, as I have seen this in some materials I have evaluated in the past.

A couple of years ago I analyzed an FR-PC/ABS painted bezel for an electronic enclosure molded in China and found that their injection molder had sworn they used the same material, but the batch that had poor performance had a bunch of silicone oil in it. This caused failing the heat performance test and very poor paint adhesion.

Like most processes, making plastic parts look like metal can be done poorly or well, and can be appropriate or not. If you want to look at some really beautiful metal coated parts, check out some of the bathroom and kitchen fixtures at your local Home Depot, Lowe's, etc. This requires specific material and design skill, but these parts are durable, wear reasonably well and look just like, particularly, chromed metal, in part because those items are chrome-plated.

As with any project, the hard part is to define as closely as possible what it is you want: Does it have to "sort of look like metal," does it have to look like a mirror, does it need to look like brushed stainless, etc., and how much abrasion, temperature resistance (e.g. will it sit on the top of a car dashboard where it could get to 180F), does the metal in the metallic look have to contribute anything other than aesthetics (e.g. conductivity)? The more time you spend closely defining the parameters, the better chance you have for success.

A little late to add anything about the difference between extrusion and injection molding, blow molding, but I want to add a practical market view including the factors of how much the molds cost and how many items are wanted.

Extrusion is a continuous process used to make sheet, pipe and profiles, film and coating, wire covering, filaments and fibers, feeding blow molders, and mixing/making pellets for other processes, including injection molding. Continuous extrusion dominates these markets -- there are ways to extrude into a closed mold, perhaps useful for low-volume applications. Almost all injection molding really does this, too, using a screw to melt the material, and then as a piston to force the melt into the mold. However, these processes are always called "injection," and the use of the phrase "extrusion molding" is confusing without further explanation.

To decide an injection mold for unpredictable demand, I would see whether more cavities will change the mold base size. If part size is small and volume is low, a 2 cavities and 4 cavities mold may use the same mold base (hence use the same machine tonnage), then it may be wise to have a 4 cavity-capable mold with only 2 cavity cut, like a bridge mold.

However, I am still a little bit doubtful on the saving we can gain by just not cutting the remaining cavity. Building a small injection mold is still the 1st preference. Then we may end up with 2 small molds, which is more expensive. However, the second mold is a duplicate mold which will not need to design, program, and go through various mold modification as the first mold. Furthermore 2 small injection molds are sometime more flexible for production planning compared to one big mold.