0
TECHNICAL PAPERS

Intelligent Modeling of Thrust Force in Drilling Process

[+] Author and Article Information
Ye Sheng1

 Accuray, Inc., 1310 Chesapeake Terrace, Sunnyvale, CA 94089ysheng@accuray.com

Masayoshi Tomizuka

Department of Mechanical Engineering, University of California, Berkeley, CA 94720

1

Address all correspondence to this author.

J. Dyn. Sys., Meas., Control 128(4), 846-855 (Feb 04, 2006) (10 pages) doi:10.1115/1.2361322 History: Received April 27, 2004; Revised February 04, 2006

In this paper, an intelligent modeling strategy for thrust force in drilling process is proposed. First of all, neural network (NN) models are developed to model the thrust force in drilling process. Second, drill head position information is included in the NN model to get better force prediction accuracy for entrance and exit drilling stages. Third, a fuzzy switching strategy is proposed to deal with the gain variation problem due to transitions from one drilling stage to another. Finally, gain variation due to drill wear is studied and the related modeling strategy is developed. Simulation and experimental results show that the proposed model works well over a wide operating range.

FIGURES IN THIS ARTICLE
<>
Copyright © 2006 by American Society of Mechanical Engineers
Your Session has timed out. Please sign back in to continue.

References

Figures

Grahic Jump Location
Figure 1

Experimental setup

Grahic Jump Location
Figure 2

Architecture of NN model with one hidden layer: (a) NNARX model; (b) NNOE model (note that the bias nodes are not shown)

Grahic Jump Location
Figure 3

APRBS signal; AL specimen, CT drill, open loop: (a) force response; (b) command feed rate

Grahic Jump Location
Figure 4

Validation results of NN models trained with APRBS signal; AL specimen, CT drill, open loop; solid lines represent the measured forces and dashed lines represent the simulated forces in (a) and (c): (a) force responses of the entire drilling process; (b) command feed rate of the entire drilling process; (c) force responses of the middle stage of the drilling process; (d) command feed rate of the middle stage of the drilling process

Grahic Jump Location
Figure 5

Different drill-specimen relationships, the shadowed area of the drill head is the part which really takes part in drilling: (a) entrance stage; (b) middle stage; (c) exit stage

Grahic Jump Location
Figure 6

Validation results of NN models; entrance stage of the drilling process; AL specimen, CT drill, open loop, solid lines represent the measured forces and dashed lines represent the simulated forces in (a) and (c): (a) force responses without the information of the drill head position; (b) command feed rate; (c) force responses with the information of the drill head position; (d) command feed rates

Grahic Jump Location
Figure 7

Fuzzy membership function for drill depth, AL specimen, H=9.3mm, EM-US=B̂S−MX=2.2mm, EM-EML=EMR-EM=MX-MXL=MXR-MX=0.8mm

Grahic Jump Location
Figure 8

Validation results of NN models; entrance stage of the drilling process; AL specimen, CT drill, open loop, solid lines represent the measured forces and dashed lines represent the simulated forces in (a): (a) force responses with the fuzzy switching strategy; (b) command feed rate

Grahic Jump Location
Figure 9

Evolution of maximum force under constant command feed rate U=3.0V, open loop, CM specimen, HSS drill

Grahic Jump Location
Figure 10

Evolution of maximum force under constant command feed rate U=1.8V, open loop, AL specimen, CT drill

Grahic Jump Location
Figure 11

Neural models with switching strategy based on SN: switches S1-1 and S1-2 will switch to the same SN channel, N is the maximum number of holes satisfying certain specifications a drill can make

Grahic Jump Location
Figure 12

Validation results of NN models; entire drilling process; AL specimen, CT drill, open loop; solid lines represent the measured forces and dashed lines represent the simulated forces in (c)–(f): (a) command feed rate with SN=1; (b) command feed rate with SN=6; (c) force responses, SN=1 for training and SN=1 for testing; (d) force responses, SN=6 for training and SN=6 for testing; (e) force responses, SN=1 for training and SN=6 for testing; (f) force responses, SN=6 for training and SN=1 for testing

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In