Web-Based Machine Vision and CMM

jabgoldfishAI and Robotics

Oct 19, 2013 (3 years and 7 months ago)

73 views

Laser Cutting Performance and Quality Evaluation via

Web
-
Based Machine Vision and CMM




Principal Investigators:
Dr.
Richard Chiou, Dr. Michael Mauk


Research Assistants:
Yueh
-
Ting Yang, Sweety Agarwal


Applied Engineering Technology

Goodwin College of Professional Studies

Fig. 1:

The configuration for laser hole cutting.


Fig. 2:

Snapshot of the Application for E
-
Machine Vision.

Conclusion

From the results, we can tell the measurement data from CMM are reliable and
accurate. The rough cutting model was built through simple vector calculation.
To predict the laser cutting quality of different material, this methodology could
be further improved for the industrial area.


However, the machine vision data need to be further investigated. From the
technical support of the machine vision system developer, we know the
accuracy of the machine vision system is about 1 mm at best. The width of the
laser cutting falls between 0.3 to 0.45 mm. This will cause the unreliable error in
the machine vision data. Higher pixel cameras and fine tuning are necessary to
to improve the accuracy to as high as 0.01 mm.



Mathematical Relation

Elements

1. Universal Laser Systems X
-
660
Laser Engraving and Cutting
Systems

2. DVT 540 C series


(color vision camera)

3. Yamaha RCX144 web
-
based


controller

4. TESA Micro
-
Hite 3D CMM

5. Workpieces to be inspected

6. Remote Server

Work pieces

cut by laser

Desired Results

Yes

No

Inspecting by

machine vision

Data from

machine vision

Compare data

Inspecting by

CMM

Data from

CMM

START

Look for

other methods

Fig. 7:

Flow Chart


Fig. 4
: Dependence of the gap width on the laser


cutting speed measured by CMM

Fig. 3:

Dependence of the gap width on the laser


output power measured by CMM


Fig. 3:

The gap width versus laser output


power measured by Machine Vision


Fig. 4:

The relation between gap width, speed,


and power.










From the experimental data we have, when power equals to 60 Watts, the
relation between gap and speed can be shown in Figure 3. We consider the
data to be linear under the condition that the speed falls between 0.3 to 1.4
mm/s, so that the average data follows the linear equation in Figure 3.



W
is the width of the gap,
V
the speed of the laser.

When speed equals to 0.3 mm/s, we can plot the result shown in Figure 4.
Consider the data to be linear that the power falls between 36 to 60 Watts. The
average data formed another linear equation in Figure 4.





P
is the laser power.

Since we consider the two equations to be linear, a plane can be determined by
the cross product of these two vectors, as shown in Figure 5.

Now we rewrite (1) and (2) into the vector form as .

Let , , then .


Take Set E to be the initial condition.
W

= 0.425 mm,
V

= 0.3 mm/s,
P

= 60
Watts.

The plane equation then be determined as (3).




Rewrite equation (3) into (3*).



The width of cutting for this acrylic material then was showed clearly.

r

z

#3 MIRROR

FOCUS LENS

#2 MIRROR

BEAM WIDOW

BEAM DIAMETER

LASER CARTRIDGE

GAS FILLED PLASMA TUBE

ELECTRODES

RF ELECTRONICS

#1 MIRROR

W

The emergence of the laser and its subsequent development
has presented the opportunity of increasing manufacturing
capabilities, as well as to expand to new areas. Laser cutting
of plastics offers advantages over conventional methods,
such as no tool wear and contact force
-
induced material
disintegration. However, the drawback in using the laser is
the cut loss uncertainty and its tolerance controllability. In the
current investigations, analytical heat conduction models are
used to determine the extent of the cut loss induced errors.
The results of the study of the effects of process parameters
on the cut loss and quality performance achieved during
laser cutting of plastic materials are presented.

Key words: Laser cut loss, E
-
quality, E
-
machine vision, web
-
based quality control

Abstract

Quality Control Methodology:


Introduction

The objective of this research is to determine the laser
cutting quality in terms of kerf loss uncertainty and
process parameters. A web
-
based quality control system
via E
-
machine vision for inspecting part dimensions and
tolerance is also described. Although the data can be
gathered for statistics in short order, the assessment and
interpretation of machine vision data is necessary; hence
a CMM (
coordinate
-
measuring machine
) is used to
validate the machine vision. Through the comparison of
the data acquired from machine vision and CMM, the
deviation of the machine vision can be estimated. Based
on the data gathered by CMM, the power and the speed
of laser cutting can also be analyzed. Thus, a model of
laser cutting for plastic materials can be developed. The
deviation of machine vision system is investigated, so
that it can be appropriately applied in industrial areas.

The main method is to evaluate the performance of
laser cutting quality through web
-
based machine
vision, to compare the data from machine vision and
CMM, and to determine laser cutting quality after
machine vision is calibrated for the laser cutting
quality control.