NIST logo
*

Measurement Science for Optimized Machining Processes

Summary:

Machining is the workhorse process that adds the most value to engineered products, since it is capable of achieving the precision required for many mechanical components at minimal process cost.  This project provides measurement science and standards necessary to meet industry needs for cost-effective, optimized machining of technology-intensive complex products made with non-conventional high-performance materials. Optimized machining requires accurate machining models and in-process measurements of process outputs (such as forces, temperatures, cutting tool wear) to continuously update these models and adaptively modify process parameters.  This project will deliver reference data sets for verification and calibration of numerical machining models including machining temperatures, deformation characteristics, tool wear rate, and a tool life for new high-priority materials needed by industry.  It will also develop an instrumented cutting tool holder to capture real-time tool wear data to construct first empirical and, later, generalized tool wear models as well as methods for process performance optimization based on these models.

Description:

Objective:

To develop and deploy measurement science to enable optimized machining of technology-intensive complex products made of non-conventional materials through combined application of validated process models, verified practical in-process measurement methods, and advanced optimization methods by 2014.

What is the new technical idea?

To enable cost-effective machining of technology-intensive complex products made of non-conventional materials, U.S. manufacturers need smart machining capability by monitoring and optimizing process performance.  This is especially important in the production of high value added products such as jet engines where new alloys are constantly being developed, and have increasingly poor machinability.  Current practice at small, medium, and large manufacturers explores machining process performance through costly and time-consuming empirical development of process recipes based on previous experience, specially designed experiments, process data analysis, and post-process product inspection. Due to the lack of better options, mostly used process data is as simple as cutting time and horsepower consumption to monitor cutting tool life.  Current standards for machining performance provide methods to empirically determine conservative processing windows that yield acceptable performance for a single combination of cutting tool and product material, where a process window consists of allowable ranges of controllable process variables.

To reach the smart machining capability, manufacturers must be able to (1) accurately estimate process outputs (such as cutting forces, power consumption, cutting tool life, integrity of machined surface, etc.) for given process parameters(cutting speed, depth of cut, feed rate, etc.), based on generalized physics-based process models, (2) use these models to identify optimum processing conditions, (3) measure process performance in real time to monitor actual process outputs, to fine-tune process models, and (4) adapt by dynamically adjusting process parameters to the optimal conditions identified using the improved (tuned) process models.

Finite tool life, due to progressive wear of the tool, is the most important process output controlling the choice of optimum processing conditions for difficult to machine materials (e.g., new titanium and nickel-based alloys, and composites).  Empirical tool life models can be developed using extensive testing, but are limited to the range of conditions covered by the tests; any change in the cutting insert, work material, or process will require a new set of tests to establish a new tool life model.  The high cost and the very narrow validity preclude determination of these models, and industry typically gets by using conditions that give acceptable tool life.  Many industries such as jet engine manufacturers do not want to risk tool failure, since in such a case their standard procedures call for the entire part to be scrapped.  However, the lack of even a rudimentary tool life model causes them to operate under conditions that are far from optimum in many cases.  In this situation, industries can use an instrumented tool holder to easily measure tool wear and develop tool life models, as well as for in-process monitoring to avoid tool failure.

Smart machining requires, generalized tool wear models, capable of predicting tool wear and tool life for a given combination of tool material and work material.  These models would relate tool wear rate to physical variables (PVs) such as contact pressure, temperature and sliding velocity, and would be independent of the tool geometry or the cutting process used.  Such a generalized wear model would be calibrated using tool wear measured under different cutting conditions that lead to a wide range of the physical variables.  A generalized tool wear model will be used to predict the optimum cutting conditions for a range of machining processes (turning, drilling, milling, etc.) so long as the tools are made of a similar tool material.

Numerical models (FEA) are needed for estimating the PVs (contact pressure, temperature and sliding velocity), for given machining process and process parameters, for use as inputs to the generalized wear models.  However, the accuracy of numerical models is limited by the accuracy of the inputs such as constitutive models for the high strain rate and high temperature behavior of the material, friction model, etc.  Therefore, they must to be calibrated based on reference data obtained by well-controlled machining tests.

The new idea is to develop and deploy measurement science solutions to enable industry partners to overcome barriers to optimized smart machining processes.  These measurement science solutions will provide: (1) the fundamental basis for physics-based modeling of machining through measurement methods and reference data sets for machining tool temperature distributions and mechanical material response to cutting, (2) the practical basis for in-process measurement of process performance through development of an industrial grade instrumented tool holder, and (3) standardized methods for process performance optimization.

Research plan:

The high degree of uncertainty in numerical machining models (due to variations in cutting tool and workpiece material properties, estimation of friction and environmental influences) necessitates (1) calibration of these models with reference data associated with the machining process, (2) continuous tuning based on real-time measurements of process outputs, and (3) dynamically adjusting process parameters to maintain optimal conditions identified by the use of improved (tuned) process models.  Therefore, a two-pronged approach will be used to enable smart machining.  While developing reference data sets for verification and calibration of numerical machining models on one hand, we will also develop an instrumented cutting tool holder to capture real-time tool wear data to construct first empirical and, later, generalized tool wear models.

Working with a consortium of industrial stakeholders, the program has identified Ti-6Al-4V as a high priority material to be investigated first.  Dramatically high tool wear rates at high cutting speed and feed limit the productivity. Using existing unique NIST facilities, the project will develop and disseminate a combination of reference data sets and state-of-the-art and practical measurement methods to the consortium and to broader audiences through technical publications. The project will demonstrate machining process metrology for the selected workpiece materials, and will correlate relevant process parameters to the measured process outputs.  The reference data will be used to validate numerical models and fine-tune inputs such as material models and friction.

In a parallel effort, an instrumented tool holder (ITH) that can measure, in real time, the tool wear rate, cutting force, and temperature will be developed to enable shop floor measurements of machining process outputs.  Empirical models will be developed for the tool wear rate and tool life as a function of machining parameters that will serve as a key input for adaptive optimization of machining processes.

Process measurements (for temperatures, tool wear rate, etc.), complemented by results of numerical simulations (such as stresses and sliding velocities) will be used to develop one generalized analytical wear model as an example.  Analytical models can be integrated into optimization algorithms for rapid decision making at the workstation level.

Since the tool wear rate is dependent on the tool and work materials, tool coatings, type of machining process, etc., it is necessary for industry to obtain data on the tool wear rate temperature, etc., for the specific tool and work combination used by them. To enable this, the instrumented tool holder will be made available to consortium members.

The project will also develop and disseminate a method for optimization of machining process performance based on these measurement methods and reference data sets. The optimization and measurement methods and systems will be (1) developed at NIST with consortium-provided input, (2) documented in draft form at NIST and released to the consortium, (3) revised based on consortium feedback and approved by the consortium, (4) implemented at NIST and validated based on unique NIST capabilities, and (5) verified through implementation at consortium facilities. The measurement methods, data sets, optimization methods and algorithms and optimized smart machining process systems will be disseminated as technical publications and provided to appropriate standards bodies as technical contributions.

Major Accomplishments:

Recent results:
  • [Output] Established Smart Machining Consortium consisting of industrial stakeholders to collaborate on machining model calibration as well as testing NIST developed instrumented cutting tool holder.

  • [Output] Recruited Prof. Vis Madhavan, a recognized expert in machining models and IR thermography, as an IPA appointment to replace Dr. Ivester, who moved to NIST Manufacturing NPO.

  • [Output] Recruited Dr. Brandon Lane as an NRC Postdoc to investigate tool temperatures in diamond turning.  He is also contributing the research developing reference data sets for titanium machining as well as developing the instrumented cutting tool holder for this project.

  • [Output] First prototype design of the instrumented tool holder was completed and presented to the consortium members for feedback.

  • [Output] In collaboration with the consortium, completed and documented the Design of Experiments for generating reference data sets for titanium machining.

  • [Output] In preparation for the machining experiments, NIST dual-spectrum videography system (MADMACS) was recalibrated.  In addition, to enable practical industrial measurements, a commercial off-the-shelf thermal camera configured and calibrated.

  • [Output] Draft technical publication detailing precise tool temperature measurement method and practical machining process behavior measurement method.

  • [Output] First draft of the reference data set of tool temperature measurements for titanium machining to be provided to the Smart Machining Consortium by the end of September.

Standards and Codes:

This project's technical area draws extensively from standards focused on other technical areas, including materials, equipment, tooling, quality, and control systems.  The project team will contribute fundamental and practical measurement methods for machining process performance and corresponding in-process monitoring and optimization methods through ASME B5 TC94 (Cutting Tools) and through the US TAG to ISO/TC29/SC2 (High Speed Steel Cutting Tools and Their Attachments). NIST is well positioned to play a leading role in these activities.

Material flow during chip formation estimated using high-speed microvideography
Material flow during chip formation estimated using high-speed microvideography.

Start Date:

October 1, 2011

Lead Organizational Unit:

el
Contact

General Information:

Alkan Donmez
301 975 6618 Telephone 

100 Bureau Drive, M/S 8220
Gaithersburg, MD 20899-8220