Design and Manufacture of a Dynamometer

Project Background:

Dynamometer Principle

Dynamometers are devices that measure force, power, or speed. They have a wide range of applications including performance testing, component tuning and efficiency testing and allow engineers to evaluate the performance of vehicles to optimize their design. Engine dynamometers measure engine parameters only, such as horsepower and torque, whereas chassis dynamometers measure these parameters at the driven wheels of the vehicle. This allows for the losses within the powertrain to be analysed as a whole and is more useful for the UCL Hypermile team, hence why the product of this project will be a chassis dynamometer. Dynamometers measure the torque and the angular rotation of their rollers and then use the following equation to calculate the power.

When this is compared to the power provided by the motor of the car being tested, the efficiency can be tested using:

Motivation

The UCL Hypermile team competes in the Shell Eco-Marathon competition in the prototype category once a year. The goal is to create a vehicle that has the highest possible fuel efficiency, whether in the battery electric, hydrogen or ICE divisions. In 2023 UCL Hypermile competed in the battery electric division, came top of the UK universities, and won a $3000 grant for Data and Telemetry. The goal of this project is to provide a working dynamometer which produces precise and accurate results so that the team can assess the power of the vehicle at the driven wheel and ensure that the changes being made to it have a positive effect on its efficiency.

Having decided on the essentiality of a chassis dynamometer, it then becomes necessary to consider the different types.

Types of Chassis Dynamometers

Eddy Current Dynamometers use Faraday’s Law of electromagnetic induction to vary the speed of a rotating shaft through the creation of resistance. The dynamometer uses the shaft of a roller as a rotor, and this applies voltage to the stator housing that surrounds the rotor[1]. Magnetic flux is then generated and due to Lenz law, the eddy current generated opposes the change in the generated magnetic field. This provides a braking effect to the shaft and an armature with a pointer is connected to the stator which measures the torque of the rotor. They are often considered to have the most precise load application however also commonly have the highest cost to create or buy due to their complexity.

Hydraulic Dynamometers use a hydraulic pump controlled by a motor to vary the resistance of the load on the rollers of the chassis dynamometer. The outlet of the pump flows through a valve whose size is dictated electronically, allowing control of the back pressure. This is directly proportional to the torque on the rollers[2]. These dynamometers are rather simpler to create than eddy current dynamometers, however, the addition of fluid requires added components such as fluid coolers, filtration systems and regular checks on the fluid to avoid contamination and degradation of the fluid.

Counter drive dynamometers[3] use a similar mechanism to hydraulic dynamometers wherein they vary the resistance of the dynamometer mechanically but instead, a motor is attached to the same shaft the roller sits on. This is controlled using a microcontroller to apply varying resistance based on the needs of the user. Their advantages include bidirectional testing and precision control however they may have the potential for synchronization issues between the motor of the UCL Hypermile vehicle and the opposing motor of the dynamometer.

Market Conditions

In the current market, dynamometers start at around £10,000 for the cheapest models and then easily extend upwards into the hundreds of thousands. This is too expensive for the UCL Hypermile team, hence this project.

One example of a cheaper dynamometer is the VT-2 Modular Dynamometer sold by Hi-Tech Performance. This dynamometer only consists of the minimum requirements, allowing the measurement of power using torque and angular rotation but without the ability to vary the load or measure other parameters. It is currently valued at £13, 410 but has a modular design so that upgrades can be bought to achieve additional testing parameters[4].

Design Requirements

Users Wants and Needs

Product Design Specification

Design Methodology

The type of chassis dynamometer that would best fit the design requirements was decided using a decision matrix.

The design was thus chosen to use a counter-drive motor to control the variable load on the driven wheel of the car.

The design task was split into three systems, mechanical, electronic and the data acquisition system. The mechanical system includes the frame and the rollers, the electronic system includes the counter-drive motor and its control, and the data acquisition system includes the sensors and their readings.

Mechanical System

The following design decisions were carried out with the safety factor kept in mind as the most important factor.

Rollers

Both one and two roller designs were considered, but it was decided that one roller had the advantage of more accurately simulating a rolling road and would cost less. The roller is to be fitted with end plates to increase the inertial load of the rollers and to provide an additional safety feature against the displacement of the car.

Frame

Various designs were conceived for the frame, but it was thought ideal to utilise the existing roller set-up and customize it to interface with the dynamometer to save on material costs for aluminium extrusion.

Support System to Allow Motor Rotation to Measure Torque

To measure the torque, the following equation is used:

Where is the torque, is the force applied and is the perpendicular distance from the pivot? It is therefore necessary that some rotation of the motor module is allowed for the torque to be measured. The gimbal setup was considered the best option, with a secondary shaft linked to the gimbal only allowing rotation independent of the roller shaft.

Connection of the DC Motor to the Roller Shaft

The direct coupling of two uncommon diameters has resulted in the required design and manufacture of a custom bracket. The other option considered was to use gears to position the DC motor beneath the plane of the rotor shaft.

Connection of Lever Arm to Gimbal

A custom bracket had been considered which would have fit over the gimbal, to which a lever arm will be welded. A different option considered was to insert a bolt through the aluminium extrusion and then use an internally threaded bar which would contact the load cell. Neither of these were chosen, the true design and its explanation are included in the data acquisition system analysis below in Figure 13.

Safety Features for the Gimbal

The completed dynamometer will stay in a workshop that is frequently used by mechanical engineering students at UCL of all years. It will likely be moved, or otherwise misused by a student unknowing of the sensitivity of the load cell and therefore quite likely to damage it. To prevent this, a spring is used in the load cell system, as shown in Figure 13, and aluminium extrusion ‘pillars’ prevent the gimbal from turning more than 5mm, thus keeping the force that can be applied to the load cell at less than its maximum rated value. The spring[1] has a spring constant of 11.96N/mm, whilst the distance the gimbal can turn through is 5mm. This means, using the equation of a spring, where is the spring constant and is the displacement, that the maximum force that can be exerted on the load cell is 59.8N. This is well within the range of the purchased load cell[2].

Electrical System

Motor Selection

The torque needed to counter-drive the roller was calculated as being around 3Nm. The most affordable and well-suited motor at around this specification was decided to be an E-bike motor[1], as commonly sold on websites such as Amazon. A DC motor was decided on, as the motor within the UCL Hypermile car is DC and if both were DC, then the Arduino Motor Shield Rev 3 can be used to extract data from both.

Emergency stop

This was considered an essential part of the system for safety, and it was decided that an excess of the shaft would protrude from the left side of the CAD model to allow for a brake system to be attached. This system will use bike brake callipers and will work as a manual emergency stop. An electrical emergency stop was deemed unnecessary due to the presence of an emergency stop on the side of the car.

Data Acquisition System

Torque

For the torque measurement three options were considered: torque transducers, strain gauges at 45° and lever arm and load cell measurement. The torque transducers were ruled out due to expense and the strain gauges were considered too difficult to accurately position so the lever arm and load cell measurement was decided upon.

The result of this is seen in Figure 13 adjacent. When not in use, the threaded nut can be moved up the threaded rod, so that even if a student were to press down on the gimbal by accident, the load cell would not be damaged. If in use, the threaded nut is moved down so that it just touches the Plain Top Spring Contact. When the gimbal rotates slightly, this compresses the spring of known spring constant, and force is applied to the load cell. The tubing ensures everything remains in place and the base plate is used to attach the top of the system to the aluminium extrusion.

Figure 13

Angular Rotation

A tachometer was considered, as was a rotary encoder and a hall effect sensor. The most accurate option due to the availability of an advanced CNC machine was considered to be the utilization of the hall effect sensor and the placement of 12 equidistant ferrous bolts in the shoulder plate of the roller.

Temperature

The lack of requirement of extreme accuracy for the temperature resulted in the most important factors being cost and reliability. This led to the choice of a NTC thermistor with results being measured by the microcontroller.

Microcontroller

As a novice programmer, it was considered that the Arduino system was more intuitive and just as versatile as competitors such as Raspberry Pi. The availability of information from the sharing of knowledge by hobbyists on forums and released by Arduino themselves is also far superior to that of their rivals.

Software

The coding language to be used to control the Arduino has been chosen to be C++, Arduino’s native coding language. The lack of experience in other coding options such as Java and Python made them less desirable than C++ due to the availability of resources for the combination of the latter and Arduino. The familiarity with MATLAB was also not considered an advantage, despite its interface ability with Arduino, as upon closer inspection the commands used to control the Arduino are not particularly like those used in previous engineering endeavours.

Progress

Almost all of the design has been finalised, however, there remain a small number of design problems to solve in relation to the motor. This is because, at the time of writing, it has not yet been delivered and it is necessary to first measure its dimensions, to ensure they align to the limited information provided by the manufacturer.

Current CAD Model

FEA analysis

An FEA analysis of the maximum load case of 500N on the centre of the shaft was conducted using Fusion 360 and the results for safety factor and displacement are shown below.

The motor was successfully tested against an electronic load using a variable resistor to ensure that the concept of the dynamometer was sound.

Products

The parts to be bought using the grant are included in the list below:

• RS Pro Silver Aluminium Profile Strut, 40 x 16 mm, 8mm Groove, 3000mm Length.
• RS Pro Silver Aluminium Profile Strut, 40 x 40 mm, 8mm Groove, 1000mm Length.
• RS Pro 6804-2RS Single Row Deep Groove Ball Bearing – Both Sides Sealed 20mm I.D, 32mm O.D.
• RS Pro Angle Bracket Connecting Component, Strut Profile 40 x 40 x 40mm, Groove Size 8mm.
• RS Pro Universal Connector Connecting Component, Strut Profile 40 mm, Groove Size 8mm.

These will all be used to help form the structural part of the dynamometer and are essential to ensure its safe running. The other parts have already been sourced, mostly from RS, excluding the metal billets required to make the custom parts and the motor.

Conclusions and Future Work

Most of the design work has now been completed and thus the manufacturing stage is beginning. The mechanical assembly is not anticipated to take too long, but the programming of the software is assumed to be likely to take a large proportion of the available time. Due to the choice of Arduino and simplified componentry, it is hoped that the programming shouldn’t be too complicated for a beginner due to the availability of information on Arduino coding online.

A Bill of Materials will be created to ensure that the manufacturing process is as organised as possible, and a user manual will also be created to ensure that future generations of UCL Hypermile are able to use the dynamometer with ease to ensure their designs are as efficient as possible.

Beyond the scope of this project are goals potentially achievable for future members of the team. These include the potential of using the Arduino to drive the car’s motor, the potential of creating a track profile for Nogaro and using it to automatically apply variable load to the car and possibly acquiring a computer solely for the use of the dynamometer.

The project will be considered a success should the dynamometer provide constructive and repeatable data that allows team members to improve the efficiency of the UCL Hypermile car.

[1] Electric e Bike Scooter Motor 24 volt 300 watt 11t chain Sprocket ZY1016(no date)Amazon.co.uk: Automotive. Available at: https://www.amazon.co.uk/dp/B01DPNMWCM?psc=1&ref=ppx_yo2ov_dt_b_product_details (Accessed: 29 February 2024).

[1] RS Pro Alloy Steel Compression Spring, 48.9mm x 15mm, 11.96n/mm(no date)RS. Available at: https://uk.rs-online.com/web/p/compression-springs/0121315?gb=s (Accessed: 29 February 2024).

[2] TE connectivity load cell, compression measure(no date)L. Available at: https://uk.rs-online.com/web/p/strain-gauges/2006113?gb=s (Accessed: 29 February 2024).

[1] Chalmers, B.J. and Dukes, B.J. (1980)High-performance eddy-current dynamometers,IEE Proceedings B (Electric Power Applications). Available at: https://digital-library.theiet.org/content/journals/10.1049/ip-b.1980.0004 (Accessed: 29 February 2024).

[2] Kohring, H.J. (2012)Design and construction of a hydrostatic dynamometer for testing a hydraulic hybrid vehicle.Available at: https://conservancy.umn.edu/handle/11299/140534 (Accessed: 29 February 2024).

[3] C. R. Hewson, G. M. Asher and M. Sumner, "Dynamometer control for emulation of mechanical loads,"Conference Record of 1998 IEEE Industry Applications Conference. Thirty-Third IAS Annual Meeting (Cat. No.98CH36242), St. Louis, MO, USA, 1998, pp. 1511-1518 vol.2, doi: 10.1109/IAS.1998.730342.

[4] VT-2 Modular Dynamometer(no date)Hi. Available at: https://hitechperformance.co.uk/vtech/vt-2 (Accessed: 18 January 2024).