I've been asked to help upgrade an old but functioning tensile testing machine (similar, but not identical to the attached photo).
Background....the equipment is used for applying a measured load to a sample to test it's behavior under load. The load is applied hydraulically, but measured entirely mechanically and displayed on the big dial shown. Historically there was clockwork chart recorder attached to the side of the equipment to provide a permanent record of the tests, this was lost years ago.
Problems.....there are two current problems, the first is that it works in imperial units, the second is that without the chart recorder there is no logging of the load data.
Proposed solution.....add some modern electronics and a display screen to display load in kgs and log all data automatically. The dial you can see in the picture interfaces to a lightweight metal rod that sticks out of the side of the equipment to connect to the (defunct) chart recorder. The position of this rod is directly proportional to the current load. So the intention is to precisely measure the position of this bar with a linear encoder. The encoder needs to be very accurate (a movement of less than 10cm corresponds to load of 10 tons) and it needs to be very low friction otherwise it can effect the displayed reading.
The current plan is use this linear actuator (here's the datasheet).
I have a number of questions firstly, is the use of a linear encoder the best approach to this problem? is the selected encoder the best option? If so, how do I interface to it? I get the impression that it might be designed only to work with the manufacturers test equipment, this is what the data sheet says about interfacing.....
Interfaces
Encoder Incremental encoders
Connection type Cable, universal, 1.5 m
Electrical data
Maximum output frequency ≤ 300 kHz
Reference signal, position 90°, electric, logically gated with A and B
Reference signal, number Electric, logically gated with A and B
Maximum load current ≤ 30 mA
Initialization time ≤ 3 ms 1) 2)
Maybe I'm missing something, but this doesn't seem to be enough information to actually interface this to some electronics.
For the curious, here's a picture of a similar tension load tester....
That encoder includes a spring, which has to be compressed or expanded by your rod. Ask which force on the rod is acceptable, and what force is required by the sensor.
Nowadays one would use strain gauges to measure a force. A search for "dynamometer" may reveal industry grade devices.
Nowadays one would use strain gauges to measure a force.
The objective is NOT to find a new way of measuring the strain, the objective is to record the strain as recorded by the existing device. The item in question is still calibrated yearly and is still used for formal testing. Despite the age the item is very relibale and doesn't need replacing
Could you fit a shaft encoder on the existing dial and read that?
That's a possibility, but the calibration authority has said that any modifications to the internal mechanism may effect whether or not it's allowable for them to issue calibration certificates.
Basically we're trying to replace the chart recorder with a modern equivalent, and we're allowed to do this because the chart recorder is external and NOT part of the formally calibrated mechanism.
What you are looking for is called a compression machine controller. They function by monitoring and controlling the hydraulic pressure supplied to the vice in the machine.
What you are looking for is called a compression machine controller.
Not quite - this machine stretches only, it doesn't squeeze.
they appear to be expensive.
hence the desire not to replace it!
Optical distance sensor.
Thanks - that looks exactly what I want. it's amazing how difficult it can be find something until you have a keyword to search for. isn't google amazing.
Are you able to recommend a particular one? preferably something that can be easily interfaced to an arduino?
the arm that sticks out probably connected to a pen on chart recorder that was mechanically wound like a clock if you take a hard drive apart there is a glass encoder connected to the read heads all bearing mounted once you take one apart you will see what i mean it determines the track info and could be connected to the arm with almost no load just a thought
rogertee:
the arm that sticks out probably connected to a pen on chart recorder that was mechanically wound like a clock if you take a hard drive apart there is a glass encoder connected to the read heads all bearing mounted once you take one apart you will see what i mean it determines the track info and could be connected to the arm with almost no load just a thought
How much is "almost"?
My guess is, there is no need to attach anything. Just point a camera at the dial. This is 2016: there is almost certainly software out there that can accept the image from the camera, locate the pointer in the image, and measure the angle of the pointer for you. The rest is just simple arithmetic.
EDIT: I just found this: Meter Reading With a Webcam
This is not exactly what you need, but at least it's a starting point.
Can you supply more precise details on the moving arm your wanting to measure.
What is the maximum stroke distance the arm will move?
How accurately do you need to measure the stroke position?
How fast will it be moving?
Possible choices might be magnetic linier encoder with linier magnetic strip as they are non contact so should not effect the arm movement much.
Another off the shelf option might be something like this but you would need to check what load is needed to move it.
Could you use a modern digital vernier gauge - one with a data output?
Possibly, I must admit I didn't realise they came with data ports. I suspect that the physical resistance provided by the vernier would be enough to effect the reading.
The magnetic linear encoders look interesting, but I can't make out much from the datasheet alone and would prefer to know more before I purchased anything.
I think trewjohn2001 suggestion of an optical distance sensor is probably best.
Something like this might be suitable, it's cheap and it's simple enough that I can understand most of the datasheet. I'm not sure that the resolution is good enough
since this is a certifiable test piece do some research into the company called Heidenhain that specialize in digital distance and position systems used on milling machines or Mitutoyo
