Load Cells – unexpected readings

Data from Load Cells is one of the most contested and questioned in Geotechnical Instrumentation and discrepancies are often blamed on the performance of the Load Cell itself.

Very rarely is it a fault with a Load Cell itself but rather issues with the system which includes the way it is mounted and its alignment and the effects of temperature on the whole installation. It should be realised that the measurement accuracy is predominately controlled by the design of the mounting devices, the quality of the installation of these devices with regard to axiality, eccentricity and changes of the ambient temperature at the measuring location.

Engineers typically expect the load from the jack used to tension anchors to be exactly the same as indicated by the load cell once the jack is removed. However, they fail to allow for some relaxation within the system and the fact that it is impossible to lock in 100% of the load from the jack.

The example below highlights such a case. Despite the load cells being calibrated at the National Physics laboratory in London they were identified as being faulty due to the fact that the data obtained from them varied significantly with that of the Jack at time of tensioning. From the site visit it was immediately clear that the load was not being locked in very well as it relied on manual tightening, by different people, using a large handmade spanner resulting in different and ineffective tightening. The design of the dome nut and seat within the load distribution plate also had varying tolerances all adding up to errors within the system.  In addition gaps between the main bearing plate and the load distribution plate were clearly visible showing that the load either hadn’t been fully locked in or that there was significant elasticity in the system.

In order to prove the performance of the load cell it was installed in line with the Jack and the rod tensioned. The loads on both corresponded demonstrating that the problem was not with the load cell but the failure of system to lock in and keep the load. The load cell was reflecting the load that it was experiencing not that which was “expected.”

It should be realised that the measurement accuracy is predominately controlled by the design of the mounting devices, the quality of the installation and changes of the ambient temperature

Unsynchronised readings

The reading taken on the jack is taken at a different time from that taken on the load cell. The strands are tensioned using the jack, the wedges are put in place and the load locked.

The strand is tensioned using the jack, the locking collets are then replaced and the anchor left to take the load. It is at this point the load on the installed permanent cell is read and does not equate the load in the jack prior to release.


The load applied by the jack diminishes during the load wedge locking process. This is because of the Young’s  modulus of the anchor and all the slight play in the fixtures and fittings of the system that cause elasticity in the system, it is the release of the stored energy in the system whilst transferring load to the anchor and so at the same time through the load cell that cause this indifference.

It is not the fault of the load cell installed but merely a part of the stressing process that is overlooked due to a lack of knowledge that trusting the instruments will help provide. In case the load is applied too fast, some accommodation time is needed for all the load to be distributed along the anchoring element.

Possible solutions

It is recommended to carry out Lift Off Testing for load cell reading versus hydraulic jack reading to identify any loss of load transfer.

Hydraulic jack longer than the squat load cell


This is more prone to inaccurate measurement of the loading due to eccentric loading causing the system more likely to contain bending moment in the system and so likely to cause friction between the ram and the seal, otherwise known as binding of the seal of the Jack. This will cause the loading to be over registered by the Jack. Unfortunately this is often understood as under registering load by the load cell.

Possible solutions

Load cells come calibrated from the factory before use onsite. The hydraulic jack ram has been used to install many anchors after its calibration, so it must be assumed that the calibration certificate corresponding to the load cell is more up-to-date than the calibration certificate for the hydraulic jack.


Hydraulic jack ram size is different from load cell diameter


This causes bending of the distribution plate which in turn causes over or under measurement of load depending on the ratio load cell / jack areas.

Possible solutions

Match the load cell and the hydraulic jack ram diameters. Usually, this is not easy to achieve as we will have to bear with the means available onsite.

Suitable, thick load distribution plates must be used to distribute the load as even as possible to the load cell – Uniformly Distributed Load UDL.

Temperature effects

Changes in environment temperature will have several effects on the readings registered by the load cell.


Thermal influences are complex to discern because it is not only the load cell that is affected but any structural elements surrounding the load cell, i.e. brick, concrete or steel structures.

Possible solutions

One way to understand the effects of temperature changes is to record the installed load cell readings together with both ambient and cell temperatures when no other changes are taking place.

An option is to take the readings always at the same time of the day, preferably early morning right before sunrise.

Considerations Possible solutions 
Are the readings alright with a different readout? If yes, then suspect low battery.
Are the load readings outside the range of the load cell? Make sure the expected loads and the load cell range match.
Is there a source of electrical noise such as a generator? Remove electrical noise source
Is there any significant temperature effect? Insulate to minimise effects
VW Load Cell: Is the correct swept frequency being used? Check swept frequency coming from readout and / or datalogger.
Strain Gauge Load Cell: Is the load cell part of an unexpected electrical loop? Check insulation resistance between the cell body and any cable. The reading should be >500 MΩ.
Hydraulic load cell – Manometer: Check for damage? Check the needle return to zero with no load; Check for leaks on the cell.
Vibrating Wire VW: Is the correct swept frequency being used? Check swept frequency coming from readout and / or datalogger.

Introduction to VW Strain Gauges

Looking at the different types of strain gauges their principle of operation, applications and tips for installation

Read More

Introduction to Vibrating Wire Piezometers

All you've ever wanted to know about Vibrating Wire piezometers including selection and installation.

Read More

Post Brexit – the new UKCA compliance

What the new UKCA (UK Conformity Assessed) mark means and when you'll be seeing it.

Read More