Analysing some plastic

I've noticed that some plastics batches are better than others. Some droop, some don't, some stink, some burn etc.

Since I want to start producing my own filament I was wondering how I could analyze and determine the composition of "good" plastic.

Would a spectroscope be good for this? (for example this one: [cgi.ebay.com] )

I doubt a visible light spectroscope will do much for you, especially as pigmentation may affect readings. Plastic analysis is done with IR spectroscopy.

---

[haveblue.org]

Heh I was just reading your blog and I got the notification about your response. Talk about coincidence

You're having the same issues as me. The "unofficial" plastic droops in the machine because it uses a higher temperature. I haven't tried using natural uncolored "unofficial" plastic but I hear pigment is a major contaminant.

Wouldn't a visible light diffraction spectroscope work as if I can manage to use an IR camera and an IR lightsource? It's still a diffraction grating after all.

Well, I'd assume you'd be able to use an IR source (as long as it covers a spectrum and is not a specific wavelength) and an IR sensitive camera (as long as it is sensitive across the entire polymer bond IR absorption spectrum from about 2500nm to 16500nm). That range is certainly way outside of what a visible light spectrometer will handle, though (you'd have to realign the lens and diffraction grating and calibrate to a known standard).

---

[haveblue.org]

Well IR spectroscopy is not as easy as shining some unknown IR source and recording an image on a receptor.
Thinking of the things you need to build a DIY IR spectrometer is somewhat humbling:
You'd have to calibrate your IR source.
You can shine this IR source, record the baseline emission, then introduce the sample, record this transmission and substract the baseline emission. Then you have to do a Fourrier analysis of this clean transmission to determine the absolute intensity of each IR frequency and this will be the spectrum you want to compare to a database of organic materials. This is not taking into account the necessary preparation of your sample (dilution or pulverisation...)

Frankly, you'd be better off going to a lab that does the analysis for you, as you don't even know if IR would be the ideal technique for polymer determination!

Just my quick 2c

Yep it's pretty clear that's hopeless.

There are literally thousands of different grades of each family of plastic. Take Polyethylene for example. From this you can get:
high density (pe-hd)
medium density (pe-md)
low density (pe-ld)
linear low density (pe-lld)
ultra high molecular weight (pe-uhmw)

Now there are grades of pe made for different processes such as roto molding, injection molding, extrusion, blow molding, casting, etc. Not only that but there will be sub grades of each that have varying characteristics like shear strength, resistance to warpage, resistance to degradation in sunlight, flame retardant, food grade, etc. What you will want to concentrate on is the extrusion grade materials. This pe will have a low mfr (material flow rate). This is basically a viscosity rating that the plastic has at a given temperature when it is forced through a specific die at a specific pressure. The higher the mfr, the "droopier" it will be. I am sure there are hundreds of mfr variances for each type of pe just for extrusion.

And thats just PE. The same goes for pp, abs, ps, pc, pa (nylon), peek, pmma, pvc, etc, etc. On a side note I strongly recommend you stay clear of PVC as it is dangerous when it is overheated - it releases hydrochloric gas and is self catalyzing when heated to degradation - ie it goes thermal nuclear. It also makes shiny metal turn rusty when that happens. Also another one to steer clear of is acetal. I believe it is commonly known as delrin. This stuff has the same amazing nuclear properties of pvc when overheated. Except it releases formaldehyde gas. Not fun. One of my process techs spent the afternoon once puking his guts out after getting a facefull of the gas. He looked in the hopper to see why the molder wasn't working properly. Damn lucky it didn't kill him.

Each time you buy feedstock I am guessing the supplier uses different materials he has on hand at the time or even uses the inhouse regrind from other runs. This is why you get different results with different batches of materials and different suppliers. What you should ask your supplier for is a data sheet on the material used. They may not want to give you that but you could ask for the trade name and product number. This way you could ensure you are receiving the same raw feedstock material "brand" from batch to batch.

I was trained in plastics processing at college about 15 years ago and went into plastic injection molding right after. I worked in the industry for about 7 years and then decided to do something entirely different (sheet metal shop). So some of the info may be a bit off as I have forgotten loads of stuff. I did however keep most of my handy processing books so when I find them I could probably help some more. I happened across this reprap stuff the other day and got interested in it. I started building an extruder today for the hell of it. Hopefully I can help you guys out (and myself once I start on my own reprap) with some different feedstock if my franken extruder works I hope this info helped.

Loren