Adventures in Building a Backyard Tensile Tester

Transcription

1 Adventures in Building a Backyard Tensile Tester International Technical Rescue Symposium November 2-5, 2017, Golden, Colorado Mike Forbes RopeCraft mike@ropecraft.org Overview The testing of rescue systems and their requisite components is an essential part of analyzing our systems for strengths and weaknesses. Many times rescuers and teams do not have access to the proper equipment for testing which results in poor or no results, unsafe test methods, or costs incurred in order to rent or hire someone to conduct the testing. In order to further knowledge and provide rescuers with a good platform by which to conduct testing, a mission was launched with the following goals: 1) Design a slow pull test bed made from easily sourced materials that is structurally sound and based on robust engineering practices 2) Provide for testing up to 44 kn with modern data acquisition technology 3) Be Portable 4) Cost less than $3000 USD. Through work with engineers, personal experience, and good old-fashioned handy work, a tensile tester was designed and built with the rescuer in mind. This paper and accompanying presentation detail the construction process and discusses issues encountered in its creation. My hope is that tensile testers will proliferate across the globe in order to advance the science of rescue technology. A set of plans will be available on the ITRS website for download after the symposium. THE DISCLAIMER - PLEASE READ FIRST! While I am no engineer nor pretend to be one, I must advise you that I can not accept any responsibility or liability in any form as a result of your use of the information contained within this paper. I have done my best to seek expert advice in its design however, you MUST consult with an engineer or other qualified individual before building this device. Many techniques used in its construction require extensive knowledge or expertise to do so properly and safely. I make no claims or guarantees to the safety of this device and cannot be held responsible for any losses, damage or injury to people or property, or losses occurring from its use.

2 Background Info I have found that the lack of readily available methods to bring systems and products to failure has been a hindrance in my rescue career. I decided to do something about this quite some time ago and have spent considerable effort in looking at what options are available. Tensile testing, tension testing, break testing, etc are all terms for applying a controlled tension to a test sample until it breaks. There are various ways this can be done: hydraulics, electromagnets, and inclined planes are just a few examples. Each test device must do the following at a bare minimum: 1. Method to pull the sample 2. A way to measure the applied force 3. Something to hold onto the test sample (grip) 4. Be stronger than the pulling device and not break when used Additional features that are common are: 1. Ability to change the speed of pull 2. Data logging of the force applied 3. Method to measure the displacement (how far you pulled the sample) 4. Control system to automate functions such as cyclic loading I wanted to build a tensile tester (TT) that would do all of the above with some key additions targeted at the rescue community. 1. Portable 2. Capable of pulling outside of the frame. Generally the pull is within the frame allowing for a very controlled test. While not as controlled, the ability to anchor the frame and pull something to failure outside of it would be very helpful in understanding a system (e.g., pulling a tree over or vehicle down the road). 4"x48" Hydraulic Cylinder 10k S-Beam Load Cell Moving Crosshead w6x20 Beam Fixed Crosshead

3 I based many parts of my design off of Sterling Rope Company s tensile tester and received assistance in the design of this tester from their R&D Director, Josh Walls. Details Reaction Frame There are 2 commons styles of TT that I have worked with, single and dual beam. The dual beam allows for a much stiffer frame while using smaller, lighter beams. The length of the frame decided upon was 18. This dimension was chosen to accommodate the hydraulic cylinder, grips, and still have enough room to place samples up to 84 long in the testing area. It is critical that the frame be significantly stronger than your pulling force or you risk bending components invalidating tests or worse, getting hurt. Calculations were based on treating the frame as a structural column and calculating the forces needed to buckle it. With that it is also possible to calculate the maximum force you can apply while maintaining low deflections and appropriate safety factors. Column Type Fa (ksi) Allowable Stress Pa (kip) Allowable Load Pc (kip) Critical Load Fc (ksi) Critical Stress Slenderness Ratio Safety Factor Column Weight W6x W6x W8x W8x A W6x20 1 beam was chosen for each rail of the TT. This is a readily available beam that is easy to work with in terms of drilling, lifting, and welding. Since the cylinder attaches to a beam spanning the 2 rails it must be sufficiently strong to support this pulling force without deflecting. A leftover portion of the rail beam was turned on its side so as to place the applied force from the cylinder along its strong axis. In order to stiffen the frame additional bracing was added in locations where it would not interfere with the function of the TT. An end plate was fashioned and a fairlead roller system installed to allow the ram to pull outside of the frame if desired. 1 Informational note: Steel beams are not described like wood beams. Beams that we typically call I beams are really referred in industry as Wide Flange Beams or W-Beams. The W6 denotes a beam that is roughly 6 deep (H) and the x20 denotes the weight per ft. The width of the beam varies based on the weight. For example, dimensionally a W6x20 is 6.2 x 6.02 (HW) and weighs 20 lb/ft. where as a W6x12 is 6.03 x 4.00 and weighs 12 lb/ft. Tables are readily available to assist with these figures.

4 Cross Heads The frame has 2 cross heads, moving and fixed, that travel along the flanges of the beams guided by rollers (stud type cam followers were used). These cross heads allow for various grips to be installed in them as well as instrumentation. The moving cross head, serves as the carriage that carries the load cell. The load cell is attached directly to the end of the hydraulic ram, then to the cross head and finally the chosen grip. This cross head is allowed to travel freely along the flanges as the ram is extended or retracted. The fixed cross head is pinned through the flanges with a grip installed in the center. The ram then pulls against the sample until the test is complete. This cross head can be positioned and pinned in locations between the flanges at 32 intervals to accommodate test samples of different lengths. For some types of tests, knowing crosshead travel distance as it relates to force is desirable. To measure force versus displacement a method of measuring travel distance is needed. A string potentiometer was built using the design created by the First Robotics Competition Team 2468 from Westlake High School in Austin, TX 2. This device converts a linear distance to rotation and outputs a small, precise resistance based on the number of rotations of the internal resistor. This resistance is measured with the same meter that is used to measure the output of the load cell. Grips The grips are what hold the samples to be tested. Since the goal is to be able to evaluate various rope/hardware/software configurations a variety of grips will be needed. Borrowing from others designs, a variable grip was built that allows for the following attachment points: 1. 4 drum and clamp (this is a standard test grip for cordage 3) 2. Steel pins of varying sizes 3. Anything that can be threaded onto a 3/4-16 UNF threaded rod. This is the attachment point for all grips on the cylinder side of the system. Hydraulic System Determining the system to use was based on balancing the needs of force generation, speed, distance pulled, and cost. The following minimum criteria were used: 1. Force >8992 lbf (40 kn) 2. Rate of travel >48 in/min (1219 mm/min) 3. Ram length of 48 in (122 cm) 4. Cost <$1500 USD In terms of force, most individual rope rescue components fail at less than 40 kn, some far less than that. This is our target force and it is easy to calculate what a ram is capable of by using various online calculators or good o l fashioned mathematics. The force a ram generates is equal to the hydraulic system pressure 2 Plans, kits, and fully assembled units available from 3 Cordage Institute International Standard CI , chapter 8.2.3, pg. 3

5 multiplied by the area the fluid is pushing against 4. In extension there is more force than in retraction. Since TTs are generally used in retraction for pulling a sample to failure, this is the measurement we are concerned with. To obtain the desired force a 4 bore cylinder was chosen with a system pressure of 1300 psi (more on this below). There are many types and styles of cylinders. This application is a relatively gentle operation for a hydraulic cylinder as the rod is only being loaded during retraction eliminating any buckling force on the rod. The cylinder also is not moving around or being loaded off-axis as they can be in construction/agricultural operations. A tie-rod cylinder was chosen as they are inexpensive, easily fixed, and readily available. The rate of travel is a function of the pump flow and ram geometry. The Cordage Institute specifies rope samples be tested at a rate of in/min ( mm/ min) 5. If you have ever participated in that type of testing it is painfully slow. Being able to test at a faster rate, especially for proof-of-concept testing and non-standards tests is essential. A target rate of 48 in/min was determined to be realistic and practical. Ram speed is controlled through an in-line flow controller on the return line. The electrical supply for the power unit became an issue as the first version utilized a 12 V dc unit that required a battery system to operate. While this leads to a very portable unit, recharging, duty-cycle, and cost made this solution prohibitive. The ac unit is still portable but requires a 240 V ac source which can be found by tapping into a dryer/welder circuit in many homes or a 30 A RV hookup if not near a residence. There are certain electrical and fluid flow calculations that govern all hydraulic systems. This power unit was designed by starting from the largest affordable electrical motor (2 hp) we could power with the available electrical service (240 V ac, single phase). The next step was to determine what flow rate we needed our pump to produce in order to move the cylinder rod at the target rate of 48 in/min. Using common fluid power formulas 6 a minimum pump flow of 2.0 gpm was needed. With these values established the target operating pressure was determined to be 1500 psi 7. In order to provide for a flexible platform to control the unit, 12 v solenoid valves were selected. This allows for easy integration with remote control units and micro-controllers. A wireless remote (2 channel, RF, 12 v, momentary switch) was selected for the initial operation with future plans to incorporate an Arduino micro-controller for automated tests such as cyclic loading. With this configuration the performance of the ram is as follows: Retraction Force: 49 kn (12252 lbf) Time: 57 seconds Extension Force: 66 kn (16336 lbf) Time: 76 seconds 4 5 Retraction Force = (π(piston diameter 2 - Rod diameter 2 )/4) x System Pressure Cordage Institute International Standard CI , chapter 8.2.2, pg. 3 6 Hydraulic Cylinder Speed = (231 x Flow Rate)/(60 x Area) or Flow Rate = (60 x Speed x Area)/231. Units: in/sec, gpm, in 2 7 Hydraulic Pressure = (1714 x Efficiency x Motor Power)/Flow Rate. Units: psi, hp, gpm.

6 Total cycle time: 133 seconds (full extension to full retraction) Force Measurement This is typically an area of great expense. I wanted to find a solution that was more reasonable. The Rock Exotica Enforcer Load Cell is an excellent solution for applications <20 kn however with a desired force application of 40+ kn this would not suffice. A common S-Beam load cell was utilized as they are inexpensive, readily available and easy to work with. A force measurement system consists of not only the load cell but a power supply and data acquisition method. The power supply needs to be very consistent as the load cell output is directly related to this input voltage. If it fluctuates then the load cell readings will also fluctuate. That data acquisition system is simply a means of measuring a small voltage accurately and extrapolating the results into a force. My ideal criteria for a system had the following components: 1. Sample rate >100 Hz 2. Bluetooth connectivity 3. Ability to data log 4. Inexpensive A device that would meet this criteria would typically be very expensive but through a miracle I discovered the Mooshimeter ( and felt that it would meet all of the above criteria for <$150. In essence it is a very accurate, bluetooth enabled, digital multimeter for your smartphone. Some of the applicable meter specifications: 2 channel simultaneous sampling (up to 8 khz sample rate) Voltage <600 V dc with >0.5% accuracy Voltage <100 mv with <15 nv/count resolution Voltage <1.2 V with <200 nv/count resolution Bluetooth 4.0 capability 24-bit resolution max >18 Effective bits at 125 samples/second Realtime graph similar to an oscilloscope The interface to this meter is through an app downloaded to your smartphone or tablet. The downside to this system currently is that you need to take the measured voltages and convert them to a force. I do this through some simple spreadsheet calculations. The app and software for the Mooshimeter is open source. Future plans are to solicit help in creating an interface that is more targeted at measuring force and being more user friendly for the rescue community. Since our target applied force is <40 kn a 10,000 lbf load cell was selected. I purchased a readily available, used power supply for the load cell online and was up and running. Initially tests were conducted by placing a calibrated Enforcer load cell inline which helped me calibrate the Mooshimeter with loads under 20 kn. Since the output of the load cell is linear (±0.03% full scale) I made the assumption that loads

7 >20 kn would be accurate; this could be a source of error and further tests are warranted comparing load cells. Portability This was accomplished simply by utilizing a 16 tandem axle trailer. The entire frame was mounted on a set of metal legs that can be either set on the ground for stationary use or mounted to the trailer for travel. The weight of the entire unit is 1250 lb. I originally considered using a smaller trailer for the frame but due to tongue weight concerns, length of the frame and it s high center of gravity a larger trailer was utilized. Component Details Structure Sourced locally from new and used steel Source: Pacific Steel & Recycling, Hydraulic Cylinder GForce Tie-Rod Hydraulic Cylinder 4 Bore x 48 stroke, Part# 11463) Source: G Force Hydraulics, Hydraulic Power Unit Chief AP Series AC Power Unit, 1300 psi, 2.06 gpm, 240 V ac., 2 hp, 1 phase Part# B Source: Bailey Hydraulics, Load Cell Optima S-Beam Load Cell, 10k lbf rated capacity, Part# OP Source: Optima Scale Manufacturing Inc., Power Supply Omega DMD-465WB Bridgesensor AC Powered Signal Conditioner Source: Meter Mooshimeter Wireless Multimeter, Part# DMM-BLE-2x01A Source: String Potentiometer String Potentiometer Kit, Part# am-2618 Source: AndyMark, Hydraulic Valve Control emylo DC 12V 2x 2 Channel 433Mhz Wireless RF Controller Source: Track Rollers KR26 26mm Cam Follower Needle Roller Stud Type Track Bearing, 10 mm Source:

8 Expenses The expenses listed below were at the time of the writing accurate and may involve some searching local scrap yards, thrift shops, or the internet for similar pricing. Grand Total $2, Item Qty Cost Total Load Cell 1 $ $ Load Cell Amplifier 1 $95.00 $95.00 Hydraulic Cylinder 1 $ $ Hydraulic Power Unit 1 $1, $1, Hydraulic Fittings/Hoses var $ $ Wiring Components var $ $ Track Rollers 16 $2.14 $34.24 Steel wide flange beam 1 $ $ Steel Plate 1 $ $ Misc Steel (angle/flats) var $62.00 $62.00 Steel Hardware (anchors/clevis) var $29.95 $29.95 String Potentiometer var $10.00 $10.00 Instrumentation 1 $ $ Nuts/Bolts var $ $ Threaded Rod 1 $19.43 $19.43 Paint 1 $35.00 $35.00 Acknowledgements I would like to thank the following people for their help with this project: Josh Walls for his gracious help with my questions regarding column buckling and shear force calculations, instrumentation, and general tensile tester design. Without Josh s help this project would have taken much longer it did. Owen Forbes for his help with drawing parts and assemblies in SolidWorks for this paper and presentation. POST STATIC FAILURE TESTING POST HEAT EXPOSURE