Roller Failure in a Clean-in-Place Bottle-Handling Line
A dairy plant's bottle-orienting machine uses a 60 mm diameter guide roller moulded from unreinforced PA66. The roller carries a steady 400 N radial belt force at 70 °C continuously, and the line is cleaned in place with 80 °C caustic solution at pH 12 for 20 minutes twice a day, plus 121 °C saturated-steam sanitisation for 30 minutes each weekend. After about six months the rollers become oval by 0.4 mm, causing bottles to jam, and a few have cracked radially near the bore. Purchasing proposes four replacements: (A) POM homopolymer at the same price, (B) PA66 with 30 % glass fibre at +20 % cost, (C) PPS with 40 % glass fibre at six times the cost, (D) bottle-grade PET at the same price. The shaft fit requires the bore to hold ±0.05 mm at 70 °C, and the roller is not a food-contact part. Maintenance records show jams begin once ovality exceeds 0.3 mm.
1. Diagnose the failure modes and explain why unreinforced PA66 is unsuitable here.
Read the worked answer
Three mechanisms act together, and the evidence distinguishes them.
Ovality is creep: a steady 400 N load at 70 °C applied for thousands of hours accumulates permanent deformation. PA66 is hygroscopic and absorbs roughly 2 % water, which plasticises it and lowers its glass transition towards room temperature, so at 70 °C in service it is well above its glass transition and its creep modulus is a fraction of the dry, room-temperature value. The ±0.05 mm bore tolerance is far tighter than the deflection the material can hold, so creep alone explains the jams.
The radial cracks near the bore point to a second mechanism: hot saturated steam at 121 °C hydrolyses polyamide, reducing molecular weight and embrittling the surface, and the caustic cleaning adds stress-cracking susceptibility at the highly stressed bore. Once embrittled, the creep-induced hoop strain at the bore is enough to initiate a radial crack.
Unreinforced PA66 therefore fails on all three counts: insufficient creep resistance at 70 °C, dimensional instability from moisture absorption and a thermal expansion coefficient of the order of 8–10 × 10⁻⁵/K, and chemical attack by steam. A ±0.05 mm tolerance on a hygroscopic, semi-crystalline polymer is itself questionable and should be re-examined.
2. Rank the four candidate materials with quantitative or mechanistic justification.
Read the worked answer
Ranking: C (PPS-GF40) > B (PA66-GF30) > A (POM) > D (PET).
C - PPS-GF40: aromatic, no hydrolysable amide or ester backbone in the same sense, resistant to hot caustic and to 121 °C steam, continuous-use temperature well above 70 °C, glass fibre giving a creep modulus several times that of unfilled PA66 and a much lower expansion coefficient. It meets the tolerance and the cleaning regime. The objections are cost (six times) and notch sensitivity, so the bore and any radii need generous design.
B - PA66-GF30: glass fibre raises stiffness and creep resistance by roughly a factor of three to four and cuts thermal expansion, so it would probably hold tolerance in dry service. However, the amide linkage is still vulnerable to 121 °C steam, and glass fibres provide wicking paths for moisture along the interface. It is the reasonable compromise only if the weekly steam step is replaced.
A - POM: excellent creep and fatigue behaviour at 70 °C and good alkali resistance, but continuous service near 90 °C is at its limit and repeated 121 °C steam exceeds what acetal copes with, with hydrolysis and property loss likely. It also machines and moulds to tight tolerance better than PA66.
D - bottle-grade PET: low crystallinity, low modulus, and an ester backbone that hydrolyses rapidly in hot caustic. It should be excluded immediately.
3. If PPS is unaffordable, what design and process changes would you propose, and how would you verify them?
Read the worked answer
Separate the load path from the polymer. Bonding or over-moulding a PA66-GF30 or POM tyre onto a stainless or aluminium core, or fitting a thin-walled steel sleeve in the bore, removes most of the creep-driving stress from the polymer and lets the metal set the bore tolerance. Alternatively increase the roller diameter and wall section to drop the bearing stress below roughly a third of the material's 1000 h creep strength at 70 °C.
Attack the environment rather than the material: replacing the 121 °C weekly steam with a 90 °C chemical sanitisation removes the hydrolysis driver and brings PA66-GF30 and POM into contention. Reducing belt tension, if the drive allows it, buys creep life directly.
If PA66-GF30 is retained, specify it conditioned to equilibrium moisture before machining the bore, and anneal moulded parts to relieve frozen-in orientation.
Verification: run a 1000 h creep test at 70 °C under 400 N with bore ovality measured at intervals and extrapolated to the six-month service interval; intersperse accelerated steam or hot-caustic cycling and re-measure; and check residual molecular weight by solution viscosity on exposed versus unexposed specimens to confirm whether hydrolysis is occurring. Acceptance should be ovality below 0.15 mm at the end of the test, giving a factor of two against the 0.3 mm jamming threshold.