Before specifying and purchasing a torque tester it is the best to fully understand the process and variables involved in torque testing. These are:

1. Applied torque
2. Sensitivity of automatic thread break torque measurement
3. Cap squeezing pressure
4. Container squeezing pressure
5. Topload
6. Speed of torque ramp, update rate of torque digitization
7. Dwell time
8. Product variations (dimension, liner variations, etc)

1. Applied torque
Generally speaking, the higher the applied torque, the higher the thread break torque. This is true up to the strip torque where the threads break/deform irreversibly in a cap tightening cycle. Without the cap/container manufacturer’s guidance, it is recommended to start with an application torque that equals to the cap outer diameter in mm divdided by two (in inch-pounds (in-lbs, lbf-in)).
The release torque / applied torque quotient depends on the specific cap design. It is usually in the 0.6-0.9 range and higher for glass and lower for plastic bottles. It is not uncommon to see values out of this range and dwell time usually decreases the quotient over time.

2. Sensitivity of automatic thread break measurement
In the automatic detection of the thread break torque, two methods may be used to validate the result.
a) Fallback based peak torque validation: this is the fastest and most cost effective way to measure thread break torque on a cap. In CR cap applications, special attention must made when fine tuning the fallback value in order to avoid validating the shell engagement as the thread break torque.

b) Rotation based peak torque validation: to overcome the problem introduced by the torque drop during the CR engagement an additional encoder can be used to validate the thread break torque. The rotation limit must be set according the worst case scenario of the shell engagement.

If fallback based validation is used during a measurement and the fallback is set at <1.5lbfin, the tester will stop in <25 degrees and display the engagement torque. To avoid the false readout, either the fallback must be increased above 1.5lbfin (the recommended fallback for CR caps is 2.5lbfin), or rotation based validation must be used and the rotation limit set at ~40 degrees.

See the topload vs. rotation and torque vs. rotation trends below to understand the variations during a CR cap removal cycle. The vertical axis represents both toque (lbfin) and topload (lbf), while the horizontal axis is the rotation in degrees.

3-5. Cap and container squeezing pressure, topload
The variation in cap/container squeezing pressure and the topload on the cap may also affect the torque reading. The pressure variation is caused by either force or contact area variation. The larger the contact area and/or the higher the force compressing the cap and the container threads, the higher the torque readout will be. Thus in some applications it is important to monitor the container and/or cap squeezing pressures and the topload force.

6. Speed of torque ramp, update rate of torque digitization
There are two phenomenons a package engineer must be aware of when setting a torque ramp setting:
a. When the torque ramp up is fast compared to the conversion time of the digitizing device, the removal torque readout on the digital machine can be considerably lower than the real peak due to the slow analog to digital conversion speed. The error originating from the low sampling speed is not to confuse with the quantization error. To understand the error originating from inappropriate digitization, look at the graphs below and/or find more information on the internet about the Nyquist-Shannon sampling theorem and resolution /quantization noise.

b. When the torque ramp up time is slower, the removal torque tends to be lower because the gradually increasing fatigue lowers the peak force required to finally break the threads. If the torque ramp is faster the thread break torque is usually higher.

Even if a manual torque tester had fast digitizer circuit, the lack of torque ramp control may still cause considerable variation from one operator to another, and even for one operator depending on how fast he/she manually applied torque on a specific package.

Analog torque signal (red) sampled with a 12 bit, 100ms AD converter. The peak readout with this digitizer is ~4.9 lbfin at 200ms.

The same analog signal (red) sampled (green) with a 12 bit, 10ms AD converter. The peak readout with this system is ~5.5 lbfin at 160ms.

7. Dwell time
In various experiments it has been established that release torque levels are highest immediately after application and then gradually decrease to a stable level over a period of time (days/weeks). The rate of the release torque decay is greatest in the first couple of hours/days and then reduces at a decreasing rate before reaching its stable level. Production processes such as hot filling or using heat activated glue cap systems can produce a big difference in the release torque readout when compared with results measured in a laboratory environment.

8. Product variations
Minor changes in mold, material and liner can be also major contributing factors in torque varations. See an example of different liner alignments and how it relates to the contact surface area and the release torque.