# Accelerometer vs pressure sensor

#### OctverQ

Joined Jul 8, 2022
12
Hi all,
Happy to have found this forum and sorry if my question has been asked before
I am a noob with very basic knowledge of how electronics work but Ive been doing some Youtube research and it got me even more interested
Basically Ive always had a passion for electronics but never had the time to get into it
I am here now because I want to make a device that simply measures the force of impact to ground(different types of surfaces, tarmac, grass, sand etc)
From youtube I got the hint that my device should include an accelerometer, I was initially thinking about a sensor/resistor so not sure which one is best for me
The accelerometer, as I understand it, measures all the movement of the device from the moment its released to the moment it hits the ground and then also bouncing back etc. until it stops.
The resistor will only sense the moment the object hits the ground and measures the force applied to it.
Isnt the accelerometer an overkill since I am only interested in measuring the impact moment?
I anticipate the device will weight max 400g
The max. height from which the device will be released will be around 2m

Joined Jan 15, 2015
7,075
Since you express the weight in grams I assume you want to measure gram force. Typically a load cell would be used where your object is dropped on a platen and the peak force is captured when your object strikes the platen of the load cell. You really don't want an accelerometer for this as I read your post. You want a load cell with a very fast response time. Load cells and force-sensitive resistors (like FlexiForce touch sensors) are the two most common methods. Really depends on exactly what you want such as sensor accuracy and other concerns. Normally sensors like this output a signal into a data recorder so you can see the actual curve and peak pressure. Load cells can be inexpensive to I can't believe it cost that much. The sensor will have a range of force so if you look at your object weight you can calculate the force it will strike a surface with based on acceleration and where it gets to a terminal velocity.

Ron

#### drjohsmith

Joined Dec 13, 2021
541
Hi all,
Happy to have found this forum and sorry if my question has been asked before
I am a noob with very basic knowledge of how electronics work but Ive been doing some Youtube research and it got me even more interested
Basically Ive always had a passion for electronics but never had the time to get into it
I am here now because I want to make a device that simply measures the force of impact to ground(different types of surfaces, tarmac, grass, sand etc)
From youtube I got the hint that my device should include an accelerometer, I was initially thinking about a sensor/resistor so not sure which one is best for me
The accelerometer, as I understand it, measures all the movement of the device from the moment its released to the moment it hits the ground and then also bouncing back etc. until it stops.
The resistor will only sense the moment the object hits the ground and measures the force applied to it.
Isnt the accelerometer an overkill since I am only interested in measuring the impact moment?
I anticipate the device will weight max 400g
The max. height from which the device will be released will be around 2m
what do you mean by
" I was initially thinking about a sensor/resistor "

#### OctverQ

Joined Jul 8, 2022
12
Since you express the weight in grams I assume you want to measure gram force. Typically a load cell would be used where your object is dropped on a platen and the peak force is captured when your object strikes the platen of the load cell. You really don't want an accelerometer for this as I read your post. You want a load cell with a very fast response time. Load cells and force-sensitive resistors (like FlexiForce touch sensors) are the two most common methods. Really depends on exactly what you want such as sensor accuracy and other concerns. Normally sensors like this output a signal into a data recorder so you can see the actual curve and peak pressure. Load cells can be inexpensive to I can't believe it cost that much. The sensor will have a range of force so if you look at your object weight you can calculate the force it will strike a surface with based on acceleration and where it gets to a terminal velocity.

Ron
Hey Ron, thanks for the reply
I am after a precise reading of the force of impact as the key here is the device to be able to tell what surface it hits depending on the measurements.(matching with pre-registered) If the device works right then it should register more or less the same reading every time it hits concrete for example, of course from the same height.
Your suggestion is interesting, I didn't know what a load cell is until now so thanks for that
The only issue I found is the design; a load cell has a specific contact surface while I was imagining my device to look more like a sphere. This is because it may rotate whilst in midair.

#### OctverQ

Joined Jul 8, 2022
12
what do you mean by
" I was initially thinking about a sensor/resistor "
Sorry I should of said pressure sensor/resistor.
there are load of models available on the market
I am not sure if these are able to measure impact or just detect the impact
like would they be able to sense slight changes in the load due to increased height for example

#### drjohsmith

Joined Dec 13, 2021
541
Sorry I should of said pressure sensor/resistor.
there are load of models available on the market
I am not sure if these are able to measure impact or just detect the impact
like would they be able to sense slight changes in the load due to increased height for example
Can you post a link to some of these please
if they have data sheet with them, and not Alb*** that would be informative,
thanks

#### nsaspook

Joined Aug 27, 2009
10,678
Hey Ron, thanks for the reply
I am after a precise reading of the force of impact as the key here is the device to be able to tell what surface it hits depending on the measurements.(matching with pre-registered) If the device works right then it should register more or less the same reading every time it hits concrete for example, of course from the same height.
Your suggestion is interesting, I didn't know what a load cell is until now so thanks for that
The only issue I found is the design; a load cell has a specific contact surface while I was imagining my device to look more like a sphere. This is because it may rotate whilst in midair.
Sounds like a good reason to use a 3-axis accelerometer as part of a 9DOF IMU sensor that will tell if the object is in free-fall, show any rotations during travel, contact and rest and give the data vectors for forces.

Joined Jan 15, 2015
7,075
Hey Ron, thanks for the reply
I am after a precise reading of the force of impact as the key here is the device to be able to tell what surface it hits depending on the measurements.(matching with pre-registered) If the device works right then it should register more or less the same reading every time it hits concrete for example, of course from the same height.
Your suggestion is interesting, I didn't know what a load cell is until now so thanks for that
The only issue I found is the design; a load cell has a specific contact surface while I was imagining my device to look more like a sphere. This is because it may rotate whilst in midair.
I believe what you are after is impact force from a falling object? Looks like this:
http://hyperphysics.phy-astr.gsu.edu/hbase/flobi.html

For reasons unknown I can't seem to post a text link so I tried this way.

Ron

#### OctverQ

Joined Jul 8, 2022
12
I believe what you are after is impact force from a falling object? Looks like this:
http://hyperphysics.phy-astr.gsu.edu/hbase/flobi.html

For reasons unknown I can't seem to post a text link so I tried this way.

Ron
I saved that page; good physics lesson right there.
I could use the formula to calculate the force of impact in my Arduino program.
This leads me to my next question. What to use to calculate lenght of travel after impact which is needed for the formula. This can be either bouncing back up(in case of concrete) or penetrating the surface (in case of sand). Does a 3axis accelerometer be able to measure this

#### drjohsmith

Joined Dec 13, 2021
541
I saved that page; good physics lesson right there.
I could use the formula to calculate the force of impact in my Arduino program.
This leads me to my next question. What to use to calculate lenght of travel after impact which is needed for the formula. This can be either bouncing back up(in case of concrete) or penetrating the surface (in case of sand). Does a 3axis accelerometer be able to measure this
How fast to you want to record updates ?
over how long a time ?

Joined Jan 15, 2015
7,075
I saved that page; good physics lesson right there.
I could use the formula to calculate the force of impact in my Arduino program.
This leads me to my next question. What to use to calculate length of travel after impact which is needed for the formula. This can be either bouncing back up(in case of concrete) or penetrating the surface (in case of sand). Does a 3axis accelerometer be able to measure this
That gets a little tricky. Been more years than I care to remember but I did some work with hardness testing which amounts to pushing a stylus into a surface and measuring the dept the stylus travels. We used the already old Rockwell Hardness Testing machines. The stylus had a given weight on it. We checked these old mechanical beast using hardness coupons of known hardness. Here is an overall paper on the subject. Twenty-Five years ago this was part of a project I was sucked into kicking and screaming. Anyway we tested our systems against samples of known hardness. Also our work involved testing steel hardness following assorted heat treating. Concrete? Sorry I haven't a clue. Thinking it would fall under Industrial or Civil engineering disciplines.

A very rough guess would be to place a stylus on a surface and zero a dial indicator (digital is fine). Next apply force to the stylus and measure travel at various amounts of force. Eventually something has to give. Again, not my forte or even close.

As force is applied it can be plotted using a good strain gauge load cell in a compression mode. Plot force and travel and just dump a few channels into a data acquisition system plotting and recording along the way. Not quite the same as dropping an object on a surface but again not my forte. Wish I could be of more help.

Ron

#### OctverQ

Joined Jul 8, 2022
12
How fast to you want to record updates ?
over how long a time ?
I am thinking the arduino to detect a drop and then transmit the measurement of the impact via bluethooth to a android program that will be able to interpret it. Then the process is repeated for every drop/measurements. Basically this is just in my head not sure if actually possible

#### OctverQ

Joined Jul 8, 2022
12
That gets a little tricky. Been more years than I care to remember but I did some work with hardness testing which amounts to pushing a stylus into a surface and measuring the dept the stylus travels. We used the already old Rockwell Hardness Testing machines. The stylus had a given weight on it. We checked these old mechanical beast using hardness coupons of known hardness. Here is an overall paper on the subject. Twenty-Five years ago this was part of a project I was sucked into kicking and screaming. Anyway we tested our systems against samples of known hardness. Also our work involved testing steel hardness following assorted heat treating. Concrete? Sorry I haven't a clue. Thinking it would fall under Industrial or Civil engineering disciplines.

A very rough guess would be to place a stylus on a surface and zero a dial indicator (digital is fine). Next apply force to the stylus and measure travel at various amounts of force. Eventually something has to give. Again, not my forte or even close.

As force is applied it can be plotted using a good strain gauge load cell in a compression mode. Plot force and travel and just dump a few channels into a data acquisition system plotting and recording along the way. Not quite the same as dropping an object on a surface but again not my forte. Wish I could be of more help.

Ron
Ron I think I know what you mean. that would be much easier to implement if we didnt have to measure from different heights.
Thanks for getting in touch man

Joined Jan 15, 2015
7,075
As to the height from which an object is dropped there are ways to do that. Depends on your hardware arrangement. Then too this takes us back to how to measure the depth of the dent.

Ron

#### drjohsmith

Joined Dec 13, 2021
541
I am thinking the arduino to detect a drop and then transmit the measurement of the impact via bluethooth to a android program that will be able to interpret it. Then the process is repeated for every drop/measurements. Basically this is just in my head not sure if actually possible
How many samples per second / minute do you need to aquire
some sensors can read every us, some every ms, some every second,
the interface they have to the Arduino will be reflected in the speed,

#### Jon Chandler

Joined Jun 12, 2008
429
A load cell for a quality scale is designed to be insensitive to the point of load – i.e., the reading will be the same whether the weight is placed at center of the scale or at the edge.

Depending on the size of the sphere (and weight of the impact surface), a small postal scale, bathroom scale or even a large industrial scale might serve as a good starting point. The desired type of scale will have a strain-gauged beam – that and the mechanical parts of the scale are the important parts. The electronics would be replaced with a Whetstone bridge circuit and ADC to capture the peak impact load.

#### OctverQ

Joined Jul 8, 2022
12
How many samples per second / minute do you need to aquire
some sensors can read every us, some every ms, some every second,
the interface they have to the Arduino will be reflected in the speed,
Good question. The impact will probably last less then 1s therefore measuring every ms would be good
us would be an overkill and probably more expensive as well?
Btw frequency here is not as important I can allow time for the data from the sensor to be broadcasted to the android.
Basically I want to allow a few minutes between every testing.

#### OctverQ

Joined Jul 8, 2022
12
A load cell for a quality scale is designed to be insensitive to the point of load – i.e., the reading will be the same whether the weight is placed at center of the scale or at the edge.

Depending on the size of the sphere (and weight of the impact surface), a small postal scale, bathroom scale or even a large industrial scale might serve as a good starting point. The desired type of scale will have a strain-gauged beam – that and the mechanical parts of the scale are the important parts. The electronics would be replaced with a Whetstone bridge circuit and ADC to capture the peak impact load.
This seems a different approach which I am definitely considering.
I would like to keep the probe (sphere) size as small as possible. Between 60 and 90 mm diameter would be ideal.
Again the probe will have to include all the electronics inside(sensors, battery, arduino, bluetooth module etc)
The impact surface I am testing will vary but I don`t have any control over it if that makes sense.

#### Jon Chandler

Joined Jun 12, 2008
429
I would like to keep the probe (sphere) size as small as possible. Between 60 and 90 mm diameter would be ideal.
Again the probe will have to include all the electronics inside(sensors, battery, arduino, bluetooth module etc)

Ahhh. I see. The problem with dropping a "probe" and measuring everything from the probe is the orientation that it impacts the test surface. A force gauge won't provide accurate, repeatable results because you can't ensure its line of action will be perpendicular to the test surface.

There, an accelerometer, or more accurately, three accelerometers (or a single tri-axial accelerometer) located at 90° to each other and sampled simultaneously will probably be required. It will take some "fancy math" to sum the output from all three accelerometers to determine the resultant acceleration and thence the force.

Instead of a sphere, could you use a cylinder with a low-friction guide to allow a unidirectional sensor? Say a round rod for the guide with a pair of screw-eyes sliding along it to keep the probe vertical to the impact surface? Then, a single accelerometer could measure the impact, greatly simplifying the measurement process.

#### drjohsmith

Joined Dec 13, 2021
541
Good question. The impact will probably last less then 1s therefore measuring every ms would be good
us would be an overkill and probably more expensive as well?
Btw frequency here is not as important I can allow time for the data from the sensor to be broadcasted to the android.
Basically I want to allow a few minutes between every testing.
Check the sensor you select,
for instance a SPI or I2C interface might be to slow
an analog one into the Arduino adc inputs might be more useful