I sign the purchase orders at a 140-person machinery company. That puts me in an odd spot: I do not choose the engineering material, but I own the budget after a mismatched part is installed and later fails. After six years of tracking invoices, emergency freight, overtime, and rework, I can tell you where most of that waste begins. It begins when we compare material names before we compare service conditions.
My position is simple: most material-related overruns are not caused by cheap materials. They are caused by incomplete specifications and late verification. And the phrase I hear most often, thermoplastic vs silicone, is a perfect example. It sounds like the right question, but it is missing the details that drive the actual cost.
The 'Nylon 6 Is Nylon 6' Mistake
We build index conveyors. For years, our dry guide blocks were made from a Nylon 6 grade that handled abrasion and load beautifully. The data sheet made sense: Nylon 6 is a tough thermoplastic with good sliding behavior and high mechanical strength, provided the operating condition matches the test condition.
Last year, an engineer used the same drawing on a new washdown station. I ordered the same Nylon 6 because the line item was proven. The part looked identical. Six weeks later, it had absorbed moisture from a hot, caustic washdown cycle and swelled enough to let the belt drive walk out of line. The guide block cost roughly $2.50. The pulleys, the belt, the alignment call, and the unscheduled downtime cost about $1,400.
No one bought a counterfeit material. No one picked the wrong quality tier. We picked the right material for the old environment and applied it to a new environment without checking. The failure was not in Nylon 6. The failure was in the phrase same as existing.
Since then, every engineering requisition has to list service conditions: temperature range, chemical exposure, moisture, UV, load type, and whether the part runs intermittently or continuously. It feels like extra friction until the first time it catches a mismatch. Then it feels like the cheapest insurance we have.
The Myth That Material Choice Is a Direct Material Comparison
The popular thermoplastic vs silicone debate frustrates me because it looks like a smart prepurchase comparison, but it skips the question that creates the invoice: what is this part actually doing?
A thermoplastic like Nylon 6 can be a strong, rigid structural part in the right environment. A silicone elastomer can do what a hard thermoplastic usually cannot: compress, seal, and keep flexible across a much wider service temperature range. Those are not two shades of the same material. They are two different families doing two different jobs.
So the question should not be thermoplastic vs silicone? It should be: what physical and chemical load will this part face? Once that is written down, the material comparison gets easier. If the part is a structural bracket near no solvents and low humidity, an engineering thermoplastic may be the sensible choice. If the part has to seal a joint against hot water, ozone, or repeated thermal cycling, a silicone elastomer is probably the responsible option, even if the unit price is higher. I have paid more for a silicone seal that lasted through four maintenance cycles than I have for a lower-cost option that failed after the first one. The nominal per-part comparison looked right. The total cost comparison was nonsense.
The thermoplastic vs silicone thinking comes from an era when catalogs were organized by material family. It is a catalog question, not an engineering or cost question.
What a Brand Name Actually Buys You
In procurement, I get asked for that DuPont plastic as if the brand name were the spec. It is not. The brand is shorthand for consistency and technical documentation. To use it correctly, you need the specific grade, not a nickname.
When I need to verify a property, I check the data sheet on the DuPont official website. The website gives more than a one-line description. It tells you the test method, the specimen condition, and the property range. A reseller's online table often forgets that half the battle.
You can pay for a premium plastic and still create a $1,400 failure if the grade's moisture-absorption rate is outside your application condition. Conversely, you can buy a less exotic material and run it for years if you got the environment right. The brand is not a magic shield. It is a trustworthy source of information, if you read it before you order.
Prevention Is Also an Installation Step
Material specification is not the only place where five minutes of checking beats five days of correcting. Installation has the same pattern.
Consider O-rings. A seal can be chemically perfect but still fail if it gets nicked during assembly, if the gland has a burr, or if the rubber is forced onto a dry shaft. Our maintenance procedure now includes a line for assembly lubrication on elastomeric seals. The common item in our carts is Permatex O-ring lubricant; it reduces friction so the seal seats into place without being cut or rolled. It costs almost nothing and prevents the most frustrating kind of failure, the one that happens after the part was correctly selected.
Good seals deserve good installation. Prevention, in other words, is an action, not a material property.
What About Over-Engineering?
The pushback I always hear is: we cannot verify every condition. We would never get products out the door. Fair. I am not asking for a lab study on every fastener. I am asking for a short functional spec: What is the temperature? What is the chemical exposure? What movement or pressure is involved? What is the expected life?
Look, I am not anti-silicone or anti-thermoplastic. I am anti-undefined service condition. If you write those conditions down, most material debates, including thermoplastic vs silicone, start to answer themselves.
You do not need more material options. You need a sharper definition of the problem. Then choose a material, verify the source, and protect the installation. The five minutes spent doing that is the least expensive insurance a production facility can buy.