Protection from shock/vibration

Hello all, I just need some suggestions on how to best protect my mega arduino from shock, vibration and acceleration...I've looked into encapsulation("potting") with an epoxy. I just don't know whether that is the easiest, simplest thing to do. Thanks for looking.

In my experience potting is a bit permanent, my projects rarely reach the stage where they are 100% finished, I always seem to want to revisit them and solder on a few more wires or dismantle them for parts. These days I normally mount my pcbs on foam blocks inside a project box. Seems to work for most applications

Cheers

How thick is the epoxy layer ?
It makes the board heavier, thus the board has to endure more bending stress due to vibrations.

Is it for vibrations (like in a helicopter) or a shock (like in a crash detection module).

If you buy an Arduino board, you have no control over the pcb copper traces or the amount of solder. If you would create your own board, you can easily make it 100 g proof.

The X-tal, the big components and the bending/vibrating of larger boards seems to me the largest problem. A ceramic resonator would be better. Avoid big capacitors and use a Mini Pro instead of the Mega board.

If I look at the Mini Pro and the Mega board, I guess that the mini pro (with ceramic resonator) can deal with a lot more g-forces. Perhaps even 10 times more.

Reducing the peak acceleration by encapsulating it in foam is the most important. A thin layer of epoxy will certainly help, but the crystal inside the X-tal component will not benefit from it.

BigusDickus:
In my experience potting is a bit permanent, my projects rarely reach the stage where they are 100% finished, I always seem to want to revisit them and solder on a few more wires or dismantle them for parts. These days I normally mount my pcbs on foam blocks inside a project box. Seems to work for most applications

Cheers

Yeah that is what I was thinking as well, applying an epoxy layer will definitely leave no room for error. But I never thought about using a protective foam...I forgot to mention that this board will be in a centrifuge with a centripetal force of 100g. And our team will have several add-ons such as: xbee shield and the xbee itself. As far as arduino choice our team has already decided on the Mega.

Best,
Cruz

A centrifuge, that is mostly continuous g-forces, and less vibrations.
Potting it completely in a block of epoxy is possible. With a centrifuge the risk of the block breaking loose and bouncing all over the place is not so big.
Attaching the shield + XBee to the Mega is a problem.
If you support the complete solder side of the Mega, there is still a lot of force on the shield and the XBee.

What about the XBee radio signal ?
You will have something simular to a fast rotation antenna.

I'm sorry, but I stick to my opinion. I think something small like the Mini Pro with ceramic resonator and with a RF module (connected with wires) would result in something that is far more capable to withstand the g-forces.

What about the batteries ? Or is there some kind of power available ?

Erdin:
A centrifuge, that is mostly continuous g-forces, and less vibrations.
Potting it completely in a block of epoxy is possible. With a centrifuge the risk of the block breaking loose and bouncing all over the place is not so big.
Attaching the shield + XBee to the Mega is a problem.
If you support the complete solder side of the Mega, there is still a lot of force on the shield and the XBee.

What about the XBee radio signal ?
You will have something simular to a fast rotation antenna.

I'm sorry, but I stick to my opinion. I think something small like the Mini Pro with ceramic resonator and with a RF module (connected with wires) would result in something that is far more capable to withstand the g-forces.

What about the batteries ? Or is there some kind of power available ?

I am not quite sure what you meant by potting it completely in a block of epoxy? Do you mean the bottom surface of the arduino itself?

I really didn't take those xbee risks into consideration, the only reason we are considering using the xbee is because we are collecting data wirelessly through Labview, we also have many inputs such as; actuators, vacuum pumps, liquid pumps, flow sensors and strain gauges.

As far as the power, the centrifuge itself provides 54 lines @2amps each and 10 lines @5amps each.

I really appreciate your input guys!

Best,
Cruz

I am not quite sure what you meant potting it completely in a block of epoxy?

Put the lot in a box and poor the potting goop in until everything is immersed. You wind up with an impervious block but it's pretty much a one-way process.

I really don't have a much of a feel for what 100G is, but a Uno weighs 30g so let's assume your shield and Xbee is similar, that's 3kg on the connectors. A lot but maybe not too bad if the boards are mounted with the connections in line with the force. However using headers for mechanical purposes is never a great idea, I would add steel standoffs if possible, unfortunately often the mounting holes on a shield don't match those on the host board. I would also support the centre of the boards to reduce flexing.

There are also the issues with crystals etc as mentioned above. Once again if the board is mounted perpendicular to the force the Gs will just tend to hold things in place, but the internals of a crystal cannot be supported. I have no experience with that.

You could always use a ceramic (as mentioned) or the internal oscillator if the crystal is a problem.


Rob

Graynomad:
I really don't have a much of a feel for what 100G is

To have 100g continuously is enormous. It is just a number so it is easy to underestimate it.
I think the Arduino Mega will fall to pieces, no matter what you do.

An Arduino isn't prohibitively expensive so why not load one with a flashing light program and try it in the centrifuge. If it doesn't survive you will know quickly to try something else before you waste time developing code and hardware.

Is it essential for the Arduino to be in the centrifuge? Perhaps only the sensors need to be?

...R

Robin2:
An Arduino isn't prohibitively expensive so why not load one with a flashing light program and try it in the centrifuge. If it doesn't survive you will know quickly to try something else before you waste time developing code and hardware.

Is it essential for the Arduino to be in the centrifuge? Perhaps only the sensors need to be?

...R

It is essential for the arduino to be on-board when the experiment is being conducted. Going back to the epoxy topic, will the epoxy be able to dissipate the heat from the arduino?

Best,
Cruz

in a centrifuge with a centripetal force of 100g

Thats a lot! I guess that figure refers to the g force at the outer extremities of the centrifuge, would it not be possible to mount the arduino as close as possible to the centre of rotation? that should reduce the actual forces experienced.

cruz_e36:
Hello all, I just need some suggestions on how to best protect my mega arduino from shock, vibration and acceleration...I've looked into encapsulation("potting") with an epoxy. I just don't know whether that is the easiest, simplest thing to do. Thanks for looking.

Adding glass cloth or glass fibers will make the epoxy far stronger. Pot inside a steel case, even stronger. Set the whole on truck motor-mounts (that take truck engine weight and vibration) and you won't have 100G + vibration issues.

Heat.... what will get even warm? Mount those parts touching on the bottom or sides of the steel case or provide a metal path to the case. This is assuming it is not 100 G's in fire?

Do you have any mechanical engineers around?

GoForSmoke:

cruz_e36:
Hello all, I just need some suggestions on how to best protect my mega arduino from shock, vibration and acceleration...I've looked into encapsulation("potting") with an epoxy. I just don't know whether that is the easiest, simplest thing to do. Thanks for looking.

Adding glass cloth or glass fibers will make the epoxy far stronger. Pot inside a steel case, even stronger. Set the whole on truck motor-mounts (that take truck engine weight and vibration) and you won't have 100G + vibration issues.

Heat.... what will get even warm? Mount those parts touching on the bottom or sides of the steel case or provide a metal path to the case. This is assuming it is not 100 G's in fire?

Do you have any mechanical engineers around?

We have initial designs to pot it inside a steel casing. But like I said we will be using the shield along with the xbee, so I'm not quite sure about protecting that from shock/vibration at 100 G's also protect the boards from flexing. I was just wondering how the heat will be dissipated since the arduino itself will be consuming power. We are a team with 2 mechanical and 2 electrical engineers. Thanks again for the input.

Best,
Cruz

If the boards flex in a composite potted block in a steel case, what -can- work? I'd be more worried about parts getting crushed if the block gets too thick. Make a base that won't bend and set the rest on top but isn't that what the case is for?

Arduino runs milliwatts. Even epoxy and glass has some heat transmission rate. Make a test! If a chip will get warm faster than that, have it close to or against a steel case wall, or put a piece of metal on the chip that also contacts the wall for a path the heat can take.