D.I.Y. Hydraulic Bridge Crusher (Templeton Secondary School, STEM Program 2022)

Our STEM class made a hydraulic bridge crusher in order to host popsicle stick bridge competitions. The bridge crusher has a Phidget Wheatstone Bridge and four 200kg Phidget load cells in order to measure the amount of the force that the hydraulic press is exerting on the popsicle stick bridge, to find out which bridge design is the strongest. It can exert 6 tons of force but can only measure up to 800kg of force, so it can only test bridges able to withstand 800kg or less. This is sufficient for almost any popsicle stick bridge competition.

Hydraulic Bottle Jack 6 ton press.

12 feet of 4x4 fir

4x8 sheet of plywood

2 20cm Tension Springs.

Roughly 150 screws, assorted nuts and bolts.

2 110cm Angle Braces.

5 10x10cm Steel plates

Phidget Wheatstone Bridge and four 200kg Phidget load cells.

Our bridge crusher is constructed of 22 cuts of plywood and fir, 8 pieces of steel (2 angle braces, 5 plates and 1 square beam), 2 tension springs, 6 nuts, 4 bolts, and roughly 150 screws. It had to be built with strength in mind, as the bridge needs to break before the crusher does. In order to achieve maximum strength for cost, we selected un-knotted 4x4 fir beams for the main structure. The main rectangular structure was supported by triangular plywood gussets in the top corners to help strengthen the structure, as well as additional plywood supports for stability. In order to ensure that the ram travels directly downwards and does not sway, welded steel guides are present. These steel guides run along the outsides of the vertical fir beams. There is a steel plate at the point where the jack meets the horizontal fir beam. This is in order to expand the surface area and reduce the pressure on the beam from the jack. At the abutments, strength is also a concern, considering that they will be supporting the bridge under load. The abutments are made of solid blocks of 4x4 fir, leveled with each other in order to provide a stable platform for the bridges. These blocks have a massive amount of compressive strength. The gap in the middle has a span of 50cm.

When under load, the load cells output a certain amount of voltage depending on how much force is applied to them.

As voltage (measured in V), and force (usually measured in N), have different units, you must convert the voltage readings from the load cells into corresponding weight values. This requires a process called calibration.

First, determine what is called the zero-level offset, which is the voltage reading of a sensor with no load/weight. Then, add a load to the cell, and record how much the voltage reading changes. This can be done as many times as necessary in order to get the most accurate trend line, the slope of which models how the sensor readings change with weight. The first formula above shows how to find the slope, or calibration value for a given sensor.

The slope, Mcal, is our calibration value, N is the number of data points chosen, V is the voltage reading, and F is the force reading. Breaking the formula into components reveals change in voltage (rise) over the change in weight (run) for every data point, and averaging all the calculated values to find an appropriate slope.

Once this calculated value has been found, it can be plugged into the 2nd formula above, which returns a weight value, F. 

Here, Vi is the current reading, and V0 is the zero-level offset.

Here is a link to Phidget's Calibration Tutorial that is more in depth.

https://www.phidgets.com/docs/Calibrating_Load_Cells#:~:text=Due%20to%20manufacturing%20variations%2C%20load,results%20in%20an%20end%20application

In order to convert our Phidget cell readings into the proper measurements that we need, and convert our readings into live-updated graphs, we wrote code that allowed us to store data and graph it, before writing over the old data with new values.

Here is a plain English explanation of what going on in the code.

Data-gathering pseudocode  

setting up (is done once)

    open a text file in write mode to allow the 

    data to be overwritten

collecting data (to be looped)

    collect data from each sensors

    multiply the data from each sensors by                          

    10 million to be easily read

    write the data to the text file

    move the writing cursor back to the      

    beginning of the text file so that the 

    previous data can be overwritten    

Graphing pseudocode

 setting up (is done once)

      create 2 different empty graphs

  graphing (to be looped)

      collect data from text files

      convert the voltage reading to force; 

      remember to increase the conversion

      factor by 10 million if the original 

      voltage level was multiply by a factor of  

      10 million

      clear your first graph

      using your 1st graph, create a time/force 

      graph using the data that has been 

      Collected

Link to project code:

https://github.com/DuctTape123/electronic-weight-code-package

Currently, the system is capable of crushing most types of bridges and displaying their maximum load on live updating graphs, which can be projected onto a large screen for viewing. Were we to continue, our next steps would include making a more versatile bridge crusher, better minimizing the calibration error, and live updating the data to a website.

Here is a link of our Bridge Crusher System working and displaying data:

https://www.youtube.com/watch?v=uRW8jOPM6MQ

We have also included our Project Report. PDF that is a lot more in depth with more explanation on everything.