Linear position Sensor for tiny Hydraulic Cylinders

[James.H]

Hi all, I hate to be the one to join a forum and start off with a question, however I've been scratching my head for so long my hair is thinning.

I'm currently working on my dissertation for my mechanical engineering degree, electronics is not my strong point, all seems a bit magical to me

Anyway, I'm trying to design a hydraulic system that will operate the inlet and exhaust valves within an engine. Currently at the point of design the hydraulic cylinders however I want to incorporate a means of measuring the displacement of the hydraulic piston, which is directly attached to the valve. Will be a single acting cylinder as keeping the coil spring to close the valve. I've searched the internet looking for a sensor type that could be build into the hydraulic cylinder as don't have much space for a side mounted option. Ideally want to keep the unit as compact as possible, the maximum valve lift will be 7.5mm. Will be run as a bench prototype so temperature isnt a huge concern, however the speed will be as hope to be able to operate as if the engine was running at up to 6000rpm. Also I'd like to be able to capture a number of readings throughout the valve lift so that I can map a lift profile to enable velocities and such to be calculated.

As far as I can see there are a few options that I have, mount something onto the piston that acts as a pickup for an LVDT sensor, this makes the total height quite large. Secondly some sort of capacitive or inductor sensor that would pick up the piston, although it seems quite a large range for these types of sensors. Thirdly would be a flow meter on the inlet of the hydraulic measuring the supply which could be translated to a lift valve.

Have included a few sketches to try and help explain what I've been babbling on about.

Realise thats alot of info for people to read on a bank holiday so very much appreciate any pointers.

 

[duke37]

Fix a magnet to the valve stem and measure the flux with a Hall effect sensor?

 

[James.H]

Trouble is there isn't really anywhere to mount a sensor due to the valve spring also the valve and spring is sunk into the cylinder head so only the top is assessable. I guess I could fit a magnet to the hydraulic piston and try postioning a sensor on the side of the cylinder. That's something I hadn't thought of will look into it more.

 

[sirch]

do you just want to know if the valve is open or closed; or do you want to know the position across the 7.5mm range. If it is position how accurately do you need to know it?

Presumably compressing the spring requires increasing pressure so could you use a pressure sensor? This could be fitted anywhere along the hydraulic line and so does not have to mount on the cylinder

 

[CocaCola]

Presumably compressing the spring requires increasing pressure so could you use a pressure sensor? This could be fitted anywhere along the hydraulic line and so does not have to mount on the cylinder

That would be my suggestion, the position of the piston should be reflective of the hydraulic pressure...

 

[duke37]

The pressure will depend on the seal friction and the inertial force as well as the spring force.

The fluid flow could be integrated to get the position.

 

[James.H]

The pressure will depend on the seal friction and the inertial force as well as the spring force.

The fluid flow could be integrated to get the position.

That had crossed my mind, I've calculated a number of various pressures that theoretically should obtain a certain lift and the real lift figures could be calibrated. My only fear is the refresh rate of a pressure sensor. If im operating at a speed equivalent to 6000rpm that only gives 0.00356 seconds to lift and close the valve. I'd much rather measure the lift directly rather than calculate it via a pressure due to the delay involved. Ideally i'd like to take a number of readings per lift cycle to be able to map a lift profile as its not just how far the valve is lifted but also the velocity its lifted at and the speed its in lowered onto its seat that I'm interested in. I know I am asking alot and thats probably why I'm struggling so much to find a solution and fits perfectly, coupled with my lack of knowledge and Im stuffed.

If there is a pressure sensor has a high enough refresh rate that would be ideal as reduces the size of the cylinder which is great for packaging the system. However Im struggling to find one. I can hook it up to a microcontroller and make the conversion sure that wont be an issue just some simple maths and probably some trial and error.

 

[sirch]

1/,00356 = 280Hz - even with 10 measurement points along the travel it is not huge from an electronics PoV. As you indicate the dynamic response of the sensor is the critical bit. There are various types of pressure sensor so some - Capacitive/Piezo - may be responsive enough. I haven't looked at pressure sensor specs but given that piezo transducers easily operate at tens of kHz I would guess that the pressure sensors are similar.

Related to pressure is strain in the supply pipe, so a strain gauge stuck on the feed pipe may do it. Strain gauges are resistive and rely on deformation of a conductor causing a variation in resistance. Presumably these would be as responsive as your A/D converter.

 

[woodchips]

I doubt a pressure sensor will give you many useful results, 280Hz is fast for fluid flow.

You have a valve spring, at one end it moves 7.5mm, at the other it doesn't move at all. So press a piezo sensor against the coil of the spring that doesn't move so far it snaps it in two. Sang here is that the coils in the middle might not be true copies of the movement at the top.

If you can get a clear view of the top of the valve stem then stick a mirror and use a reflective infra red sensor. Probably have to make one from a separate transmitter and receiver rather than an all in one, but will easily work to those frequencies and also light enough not to affect the valve itself.

Bob

 

[James.H]

Is there any type of sensor that could mount to the outside of the hydraulic cylinder and detect the piston through the cylinder wall. Due to the design of the cylinder head the valve spring is sunk right into the cylinder head making it very difficult to attach anything to or route wires.

This is the sort of thing I'm thinking of.