Grind-Trak Development and Implementation

System Approach to Grinding

System Approach to Grinding


Grind Trak is a portable diagnostic tool for in-process monitoring of the grinding process. It enables the capture of various signals (currently, power and displacement) which serve as a signature of the grinding process.
These signals give an insight into the fundamental microscopic interactions during the manufacturing processes, which in turn help to establish the cause and effect relationship between the input, process and the outputs of the manufacturing process viewed as a system. Frequently the signals and their observed variations by themselves are adequate to identify opportunities for solving the process problems such as size holding, burn, chatter, etc. in the case of grinding processes. In addition to such signature analysis – called “Feature Recognition” – the signals can be used for process improvement, process auditing, etc. Such intimate knowledge of the manufacturing processes are also the seeds for predictive or deterministic innovation of novel manufacturing processes. The work reported here was carried out to demonstrate these wide variety of possibilities in one specific manufacturing process – the abrasive grinding processes. The work reported here is preliminary and lays the foundation for long-term and sustainable developments in grinding process innovation and their industrial realization.

Grinding process signals have been measured and, analyzed for decades. Few attempts exist across the globe for a portable device. Development of such a tool in a university setting, and extending its use for practically relevant industrial applications has been one of the key features of this work.

Traditionally in industries, grinding process is considered more of an art than science – and the skilled operator a wizard. The major reason for this being the transient nature of the abrasive interactions – due to a large number of influencing factors such as wheel, dresser, component geometry and work material, coolant, process parameters, machine tool characteristics and the variation in all these control parameters through time – and the quality and productivity requirements which make the process amenable only to a skilled operator. To handle complexity as a result – there is a need for an approach that organizes all this information as an Input-Transformation-Output system. This method has been developed and is called the “System Approach”. Such an organized method also allows us to focus on the transformation and details behind it – “the science behind the process” – for which we need process signals that are collected in-situ from the process. Acquiring various process signals allows us to study this science in an industrial setting. By understanding the science behind the process and by using diagnostic tools to acquire, monitor and analyse in-process signatures, we are able to:

  • Solve process related issues
  • Optimize the process
  • Audit the system, compare with a benchmark.

The mobile nature of this diagnostic tool and its use enables any of the production operations where it is used, to be treated under controlled conditions. Otherwise such scientific treatment is possible only in a few laboratory equipment available in a few universities or in the R&D centers of large companies. When solving process related issues, the major benefit of such an approach is the scientific basis as to what the root cause is, for a given issue with the grinding process enabling us to find global, long-lasting solutions to the problems rather than a local or transient solution. Furthermore, it enables digital storage of the current status of a grinding system – similar to an ECG signature for humans. The use of such digital data of in-process signals from real life manufacturing processes and their coupling to big data analysis is envisioned as a wide open field for innovation in the manufacturing sector.



As a part of a project at IIT Madras – on development of a high precision grinding machine tool, a project being led by Prof Ramesh Babu, an understanding was required of the grinding process and how the grinding machine tool impacts the grinding process and hence the performance of grinding operation. To know more about the grinding process and the impact of various input parameters (machine tool, abrasives, coolant, work material, process parameters, and other tooling) on the grinding process and hence the outputs (quality, cost, and cycle time), we needed a way to monitor and record the grinding process in machine tools used under industrial production operations using digital diagnostic tools.
I worked on the development of such a tool, the guidance for which was given by – by Dr K (Subbu) Subramanian of the STIMS Institute. Under the guidance of Dr Subbu, we learnt about the grinding system – how to view it as an Input Transformation Output (ITO) system. We also learnt about the six fundamental microscopic interactions in the grinding process, how each of the input parameters affect the process and what the outputs are. We also learnt that by studying certain in-process signals as a signature of the process, we can determine certain critical features taking place in the process therefore helping us to get one more data point of the grinding process, in addition to the result on the work-piece.

Project Overview

Project Overview


Project Definition and Organization

The project consisted of four major parts:

  1. Development of Grind-Trak
  2. Deployment in Univeristy – as part of the Next Generation Precision Grinding Machine Tool Development (NGPG) project at IIT Madras
  3. Implementation into the mainstream industry practices

Development of Grind-Trak

Since the device is intended for use in broad industrial setting – it had to meet certain requirements:

  • Portable – the unit must be light weight, rugged and easy to carry.
  • Short deployment time – the unit must be capable of being installed on any machine in less than 30 minutes.
  • Modular – since more sensors will be added in the future, the kit should provide a provision for that.
  • High sampling rate – to enable real-time signature of the grinding signal to capture the grinding process phenomenon with very short transient time of the order of milli-seconds and smaller.

Implementation in university

As a part of the project to develop the next generation grinding machine tool we needed to be able to:

  • Replicate the industrial grinding process conditions (equivalent diameter, specific material removal rate, chip thickness, etc.) in a laboratory setting.
  • Be able to extract relevant features of the process that help evaluate the machine tool performance such as dynamic stiffness, maximum possible material removal rate, etc.
  • A report of the same can be downloaded here [glyphicon type=”download”] .


Implementation into mainstream industry practices

Grind Trak case study

Grind Trak case study

Two modules of the Grind Trak has been developed The tool is being actively used to get the process signature in a wide variety of precision grinding process applications.

Case study presentation [glyphicon type=”download”]


The tool could be used in the following ways:

  • Document the process signature for each new machine being shipped (Benchmarking)
  • As a tool to assist the Application team in process problem solving and process optimization (Troubleshooting/Optimization)
  • As an educational tool, to teach the grinding process (Education)

As a part of this process, there was also a requirement to development of additional toolkits to augment the scientific approach to grinding. These toolkits help to make important calculations of the process parameters quickly of a wide range of cases.


Grind Trak Schematic

Grind Trak Schematic

Major areas of learning:

  • Development of Grind Trak
      • LabVIEW
        • Adapted the skills required to extensively use LabVIEW
      • Digital Signal Processing
        • Conditioning of signals received from the sensors
        • Signature analysis and feature extraction
        • Peak detection and trend analysis
      • Sensors and Instrumentation
        • Evaluate various options available for sensors – power, displacement.
        • Evaluate sensors available from across the globe – UK, USA, Japan, etc.
        • Asses the characteristics of various sensors and make a choice of cost vs performance.
      • Assembling of the kit and data representation
        • Developed a Graphical User Interface (GUI) with various modules for logging, analysis and presentation of signals
        • Assembled all the tools involved in the form of a portable kit which can be easy relocated to any machine and deployed within a span of 30 minutes.
  • Use of the Grind Trak – implementation , evaluation
    • System approach to grinding (Dr Subbu’s crash course at IIT)
    • Science behind the grinding process (Dr Subbu’s course, various books)
      • STIMS Institute
      • Handbook of machining with grinding wheels, Marinescu et. al.
      • Grinding Technology, Malkin and Guo
      • Various resources on the Internet
    • Process documentation
      • Developed a process documentation sheet, which acts as a checklist to capture most of the important aspects of a grinding system to enable efficient first-time capturing of the most important grinding system information.
      • Developed a Knowledge Integration Document that captures and represents the system information, the process signals, analysis and the technical and the system outputs for a grinding process under study.
      • Based on Dr Subbu’s course and subsequent guidance.
    • Signal Analysis and interpretation (Dr Subbu’s course, his guidance working with the Grind Trak)
      • Learnt the features to analyse in a captured signal
      • How these features represent the microscopic interactions taking place and hence the quality on the component?
      • Pattern recognition, trend analysis – for e.g. gradual increase in the peak power during grinding after dressing
      • Derived analysis – Power vs. Material Removal Rate (MRR) and its implication on the grinding process
    • Development of toolkits for calculating the process parameters and the cost per component

Most of the interactions with Dr Subbu were online –through such online interactions and personal discussions was a good distance learning exercise. Apart from this, the other sources of information in each of the cases above were the internet and my past experience of working on various projects at the Centre for Innovation (CFI) at IIT Madras.

Grind Trak kit

Grind Trak kit


A modular unit has been assembled while meeting the space and cost constraints

  • The unit was beta tested in 2-3 industrial settings.
  • Based on the feedback, improvements were made and then extensive testing was conducted in an industrial setting at Micromatic Grinding Technologies Limited (MGTL), having been used successfully in more than 40 cases.

The unit has been used gainfully in an industrial environment:

  • Learnt about the System Approach and applied this knowledge to solve process related issues like burn, poor surface roughness, chatter, etc. thereby validating the use of Grind Trak as a process solving tool.
  • Used this tool for process optimization – primarily reduction of cycle time and increase of dressing skip.
  • Used this tool as a benchmarking tool – for comparing performance of different wheels, identifying the difference in the process behaviour on two machines grinding the same component.
  • Established a protocol whereby, the baseline process signal of each machine is captured for every machine.
  • Provided training at both plant locations to the application engineers on the grinding system, and the use of Grind Trak and analysis of process data and its documentation.

Also used this as a lab exercise and teaching tool for Dr Subbu’s Grinding Technology Workshops.

Lessons Learnt


  • Selection of sensors based on cost and objective requirements from suppliers selected across the globe.
  • Development of user-friendly Graphical User Interface (GUI).
  • LabVIEW application development
  • Working knowledge about the grinding process.
  • Analysis of signals – feature and pattern recognition.


  • Working within the constraints of time, cost and people
  • Leading an effort – in implementing the use Grind Trak for process documentation and process problem solving
  • Working with cross functional teams with a common purpose of learning about the grinding system and its implementation.
  • Ability to work across many worldwide suppliers – securing information from a wide variety of sources.
  • Distance learning skills – learning about the grinding process through distance courses conducted over Skype by Dr. Subbu.

Key achievements:

  • Initiative –
    • Designing and building something which may be of relevance to the users of a manufacturing process
    • Showing the industry, and helping them use it gainfully, without fear of them thinking it is useless
  • Result orientation –
    • The grind-trak development and implementation was carried out from the idea stage to a widely accepted commercially viable unit in less than 18 months!
    • This required substantial perseverance and navigation through an industry sector (manufacturing) that is deeply rooted in traditional ways and not eager to make major changes in short time duration.
    • The risk involved in changes in the manufacturing operations is high, unless the solution has high record of success with clearly  identifiable benefits.
  • Followed the model of prototyping and learning in each step and building up on it.
    • At each stage, the Grind Trak was unit tested, each of its parts (hardware and software) individually tested with real world examples to ensure their effectiveness before being integrated and deployed.

Future scope of this project:

  • Use as a deployment in grinding machine tools to provide real-time process data to the operator to enable better control of the process and the machine tool.
  • One of the basic diagnostic tools used in the process diagnostics for Grinding Process Solutions (GPS) – a new business initiative of MGTL and the STIMS Institute, dedicated to grinding system optimization, process problem solving and auditing.
  • Addition of more sensors such as vibration and acoustic emission to get additional unique information about the grinding process are being explored.
  • Development of an application knowledge base which can be used for design and analysis of grinding cycles.
  • Development of an intelligent system which predicts the health of a grinding system and predicts the correct process parameters to be used for any given application.
  • As a control system which can take corrective action based on system drifts and brings the system to its most optimal state.
  • Big data analytics to gain insights into the operating conditions, correlations between the inputs – process – outputs of a grinding system, usage, load and performance aspects of a machine tool.