Measuring ice depth at a hockey rink. What measurement tool should I use?

[huttonstephen]

I work at an ice hockey rink driving a Zamboni. As part of my job, I keep the ice thickness to within a specific range (1.25 - 1.5 inches thick). It is a challenging job with many environmental factors working together that cause the ice thickness to fluctuate over time. Small variations in ice thickness can drastically affect the amount of energy required to keep the ice at the optimal temperature for ice sports.

I am looking to build a tool using Arduino to help me monitor several environmental conditions in the arena, but the one I am looking for help with today is in measuring the ice thickness. I have found that there is a sonar sensor that can measure proximity, but I am not sure that would be the right tool for my use case. Ideally I would like to be able to attach the measurement device to the Zamboni machine and have the ice depth measured in real time as I clean the ice.

There is a tool that is in use already in the market, and in fact I use it on my Zamboni. It is a laser leveling system that features a laser projected as a plane 15 feet above the ice surface. A receiver is attached to the zamboni on a tall pole that is then used to measure the ice thickness and in turn controls the blade cut depth. This is an active system that I do not want to interact with or use for a few reasons. I am doing this arduino project as a hobby project, and the leveling system is a third party proprietary system that I can't collect the data from. The ice thickness data collected from the laser level system is not collected- it only displays real-time the thicknesses as the zamboni drives over different areas.

I want to be able to store the ice thickness data and use it in conjunction with other environmental data from the air, water and ice. I am trying to create a complete tracking system that can be used to better understand the optimal conditions for ice making and to take the guesswork out of things.

What tools or sensors exist out there that could be used to measure the thickness of the ice?

I am able to bury different objects in the ice on the subfloor, for example to create a grid or array of sensory tools.

[xfpd]

@huttonstephen

measure the thickness of the ice

Is the flatness of the poured material below the ice level within 0.25 inch? The water on top will level the skating surface regardless of the quality of the subfloor. Map fresh ice for inconsistencies. Replicate a vernier caliper's depth gauge with a "hot poker" center pin as a depth gauge and a 12" diameter, round, STEEL base to map every square foot (hhahahhha!)

[Paul_KD7HB]

A laser time-of-flight distance sensor might be modified as a first trial. Measure at an angle and sense the pulse of the reflection from the top surface of the ice. At the same time sense the reflection from the bottom of the ice. Compute the ice thickness based on the time difference of the two reflection pulses.
Are you willing to do research or want others to do the research for you?

[huttonstephen]

@Paul_KD7HB

A laser time-of-flight distance sensor might be modified as a first trial. Measure at an angle and sense the pulse of the reflection from the top surface of the ice. At the same time sense the reflection from the bottom of the ice. Compute the ice thickness based on the time difference of the two reflection pulses.
Are you willing to do research or want others to do the research for you?

Was there something in my post that indicated that I am unwilling to do research? Sorry if that is how I came across! The opposite is true! I am a novice with electronics and Arduino, so while my curiosity is high my knowledge is not, so please forgive me for that.

Im curious what sensor and type of material you are thinking about that could accomplish what you’ve suggested?

[huttonstephen]

@xfpd

@huttonstephen

measure the thickness of the ice

Is the flatness of the poured material below the ice level within 0.25 inch? The water on top will level the skating surface regardless of the quality of the subfloor. Map fresh ice for inconsistencies. Replicate a vernier caliper's depth gauge with a "hot poker" center pin as a depth gauge and a 12" diameter, round, STEEL base to map every square foot (hhahahhha!)

Yes, the flatness of the floor is fairly accurate (to within 1/4” across the subfloor) and consistent. It is a new rink, having just been built 4 years ago.

The water does not lay perfectly flat unfortunately. The problem is that the zamboni machine drives a pattern around the rink that takes several minutes to compete. By the time the Zamboni has almost completed the pattern, the water that was dispersed at the beginning of the pattern has frozen, whereas the water dispersed at the end of the pattern is still liquid. This causes a slight unevenness in the thickness, which builds every time the ice is cleaned.

There is also the issue where the zamboni tends to be driven slower along the outside perimeter of the ice so as not to collide with the dasher boards. A skilled operator can reduce the ice thickness unevenness due to the change in speed along the boards, but it is never perfect, and so the ice builds up more quickly there too.

[qubits-us]

Hi @huttonstephen ,

Welcome to the forum..

Interesting topic..
found this..
The Use of SONAR to Measure Ice Thickness
looks like there's been some studies..

good luck.. ~q

[xfpd]

@huttonstephen

water that was dispersed at the beginning of the pattern has frozen, whereas the water dispersed at the end of the pattern is still liquid

Start with colder water and slowly increase the temperature (I think that is how water on ice works).

@huttonstephen

zamboni machine drives a pattern around the rink

I am familiar. Former season ticket holder. ZAMBONI - All upper case per registered trade mark. Heh.

Have a commercialbreak, shovelgirl (for the pennies and slush) run an "edger" hand-held, tinyZ on the dashers?

I would expect the arena to designed and execute all the measuring, then send to the ZAMBONI the messages to change speed, water temperature, blade height... and work the ice jam poker....

[Paul_KD7HB]

@huttonstephen

Im curious what sensor and type of material you are thinking about that could accomplish what you’ve suggested?

Search for laser distance sensor. They will not work for detecting separate reflections, but will give you the idea of how to make the measurements.

[dougp]

@huttonstephen

I want to be able to store the ice thickness data

Then is it needed to know where a particular height measurement is located on the rink, ie. 2D map?

@huttonstephen

The ice thickness data collected from the laser level system is not collected- it only displays real-time the thicknesses as the zamboni drives over different areas.

Kind of out there but could you build something to observe the 3rd party system's display and record its measurements indirectly?

@huttonstephen

the zamboni machine drives a pattern around the rink

Same pattern every time?

If you could know the ice thickness a few feet ahead you might use @xfpd's idea and adjust water temp. on the fly.

[xfpd]

@dougp

Same pattern every time?

Depending on stadium and number of vehicles.

1. Down the center to the far end
2. Right turn against the boards (ZAMBONI seat is on left, rear)
3. After the perimeter is complete, overlap the first, center path to the end boards
4. Overlap newest LEFT side path over the center line path
5. Overlap newest CENTER path all the way to the end boards
   This pattern keeps the turn radius the same through all turns.
   
   

Listen to the Gear Daddies

[DaveX]

@huttonstephen

Yes, the flatness of the floor is fairly accurate (to within 1/4” across the subfloor) and consistent. It is a new rink, having just been built 4 years ago.

Since the bottom is level to within 1/4", most of the ice depth is controlled by the top surface. Sensing your distance down from the existing laser system seems most straightforward. Heck, maybe an arduino could drive a light-sensor array up and down vertically with a stepper to track the height of the laser level above the machine, and record that data.

[kmin]

Is optical sensing really the best approach for this? I would imagine that dirt, bubbles, crystal structure, scratches etc have quite much effect on measurements.
What about some capacitance/impedance/frequency change? Recalibrate a pinless non-destructive moisture meter used for building materials?

[dccontrarian]

Based on what you said about the commercially available system, I don't think thickness of the ice is really what you're after, it sounds like flatness is the real goal. Since the substrate is reasonably flat thickness will be fairy consistent if the ice is flat, but +/- 1/4".

@huttonstephen

There is a tool that is in use already in the market, and in fact I use it on my Zamboni. It is a laser leveling system that features a laser projected as a plane 15 feet above the ice surface. A receiver is attached to the zamboni on a tall pole that is then used to measure the ice thickness and in turn controls the blade cut depth. This is an active system that I do not want to interact with or use for a few reasons

My first recommendation would be to see if this system has an output and figure out how to read it. Second recommendation would be to use the laser signal and make your own receiver. If you are successful you can move to using an off-the-shelf laser level, they cost maybe $150 for a good one.

Typically with a system like this you only care if you're too high or too low. So you make a rod, and you put three optical sensors on it, with the middle one slightly overlapping the top and bottom ones. When the middle sensor is lit, you're OK. If either the top or bottom is lit you're out of spec and you know which direction to move. If none of the sensors are lit you're way out of spec, and whichever sensor was last lit tells you which direction you're off. If you need more granularity you can add more sensors.

[dccontrarian]

If you need to know your location as you drive around logging thickness, it's a little more complicated but doable.

If you put three lasers around the rink at known locations, and construct your sensor so that it only reads when it's pointed directly at the laser, if you spin the sensor and record the angle where it's lit you can triangulate your position. The more lasers you have the more accurate the position.

You could also have a spinning laser that moves around the rink and a network of sensors around the perimeter.

[xfpd]

@dccontrarian

I don't think thickness of the ice is really what you're after

It is.

@huttonstephen

ice thickness can drastically affect the amount of energy required to keep the ice at the optimal temperature

[huttonstephen]

[quote="xfpd, post:7, topic:1265612, full:true"]

Start with colder water and slowly increase the temperature (I think that is how water on ice works).

Unfortunately that is not how the Zamboni machine works. Water is heated one time before the Zamboni drives onto the ice to begin working. It cannot change the temperature of the water on demand, and certainly not to the precision that would be needed to accomplish that idea. Though that is a fascinating idea. I have heard of one commercially available tools which aim to give operators fine tune control and automation of the flow rate of the water, I don't think there have been any commercial attempts to precisely control the water temperature at moment of contact with the ice surface. And for my personal situation, where all of my existing tools need to remain in production, I can't modify my zamboni machine to test something like that, nor do I have the budget.

Have a commercialbreak, shovelgirl (for the pennies and slush) run an "edger" hand-held, tinyZ on the dashers?

I would expect the arena to designed and execute all the measuring, then send to the ZAMBONI the messages to change speed, water temperature, blade height... and work the ice jam poker....

Typically the zamboni machine, and in the case of the one that I operate, is controlled manually by the operator. The speed, the water flow rate, water temperature, and blade cut depth are all controlled by the operator. On my machine, there is an automated blade depth control tool that uses a laser level system. It doesn't work very effectively, as the readings are inaccurate, the cut is imprecise, and the system settings are such that the maximum variation of cut depth is not enough to overcome the other environmental factors that cause the ice depth to increase. In other words, the laser level cut system can't keep the ice cut down to the level that it needs to be, and the ice gends up getting too high in certain areas of the rink.

The nature of the tool is this- the Zamboni's cutting blade is 72 inches wide, and it needs to be set perfectly in line with the conditioner's runners (that flat metal bits that glide along the ice surface at the back). Imagine the blade as a giant ruler that spans the entire width of the ice it is cutting. If you picture a long ruler lying flat on a table, both ends should touch the table equally. If one end is raised or lowered, it won't be flat anymore. The Zamboni blade works the same way. Both ends of the blade must be level with each other to cut the ice evenly. Except that this concept performs the best when the ice itself is flat, which it isn't! the ice takes on a slight bowl shape where the outside around the boards and in the corners of the ice are raised and the middle of the ice sheet is more flat.

[huttonstephen]

Then is it needed to know where a particular height measurement is located on the rink, ie. 2D map?

That's right- I'd like to take these ice depth measurements and create a 2d map from the data. Used in conjunction with air temperature and humidity, ice temperature data, energy consumption, and refrigerant temperatures, I am going to make a tool that can help me spot "problem areas"- for example if there is a slow refrigerant leak, or damage to the subfloor, or an issue with outside air drafting into the arena, or some other unknown issue, these would become easier to diagnose with environmental data being collected semi-autonomously.

Kind of out there but could you build something to observe the 3rd party system's display and record its measurements indirectly?

That is a really clever hack that could work! the laser level system has a small display onboard the zamboni and it updates the readings a few times per second.

Same pattern every time?

yes, the pattern is the one that @xfpd posted.

If you could know the ice thickness a few feet ahead

That is the problem! the laser level system knows the ice thickness right here, and blade adjustments are made in response to that measurement, but since the zamboni is driving forwards there is lag between the system recognizing that the ice is too thick and the blade itself being adjusted. When the blade should be cutting deep, the zamboni may have moved to to an area of the surface where it doesn't need a deep cut anymore!

I am able to manually set the blade depth, and I have time set aside every day to do a slow 'ice maintenance' session- I can drive any pattern and I have a few other hand tools and cutting instruments that I can put to work. Once I create a 2d map for myself of the ice depth, my work can become more targeted and efficient.

[DaveX]

@huttonstephen

It doesn't work very effectively, as the readings are inaccurate, the cut is imprecise, and the system settings are such that the maximum variation of cut depth is not enough to overcome the other environmental factors that cause the ice depth to increase. In other words, the laser level cut system can't keep the ice cut down to the level that it needs to be, and the ice gends up getting too high in certain areas of the rink.

How high is a too-high problem area? How accurate are the laser measurements? How much is the maximum variation of cut depth? How much can the ZAMBONI change the ice depth?

I'd think that if you get a record of the system's measurements, or independently measure the height of the laser over the machine, you should be able to get a fresh 2D map each session.

Since the pattern is the same, you might be able to use time, (or distance? wheel revolutions?) or time+magnetometer/compass, to map the heights recorded during a session onto a 2D trouble map. And then use the 2d trouble map, or maybe just a replay of the timeline (distance-line) series of measurements to give the operator the "the ice thickness a few feet ahead" during the last run.

[dccontrarian]

@huttonstephen

That's right- I'd like to take these ice depth measurements and create a 2d map from the data. Used in conjunction with air temperature and humidity, ice temperature data, energy consumption, and refrigerant temperatures, I am going to make a tool that can help me spot "problem areas"- for example if there is a slow refrigerant leak, or damage to the subfloor, or an issue with outside air drafting into the arena, or some other unknown issue, these would become easier to diagnose with environmental data being collected semi-autonomously.

So just to clarify: which is more important, that the ice be consistent in thickness, or it be completely flat? Ideally, should it follow the imperfections of the subfloor perfectly at the same thickness throughout, or should it be flat on the surface and have the thickness vary somewhat to account for the imperfections in the subfloor?

[dccontrarian]

@huttonstephen

That is the problem! the laser level system knows the ice thickness right here, and blade adjustments are made in response to that measurement, but since the zamboni is driving forwards there is lag between the system recognizing that the ice is too thick and the blade itself being adjusted. When the blade should be cutting deep, the zamboni may have moved to to an area of the surface where it doesn't need a deep cut anymore!

It sounds like the answer is to measure out in front of the Zamboni, and predict where the blade will need to be in a few moments. Then use the speed of the machine and what you know about how fast the blade can be adjusted to get the blade where it needs to be when it gets there.