the answer is blowin’ in the wind - admin.eoh.co.zaadmin.eoh.co.za/content/nasa_ames.pdf · fall...
TRANSCRIPT
23FALL 20 03• h o t l i n k s
A E R O D Y N A M I C R E S E A R C H
THE ANSWER IS BLOWIN’IN THE WIND
InTouch HMI MonitorsWind Tunnel Use at NASA Ames ResearchCenter.
T H E N AT I O N A L A E R O N A U T I C S & S PA C E A D M I N I S T R AT I O N
(NASA) I S O N E O F T H E WO R L D’S L E A D I N G R E S E A RC H I N S T I-T U T I O N S S T U D Y I N G A E R O D Y N A M I C P H E N O M E N A . Much of the
advanced work conducted in this subject area is performed in the Fluid
Mechanics Laboratory (FML) at NASA Ames Research Center in Mountain
View, Calif., using in-draft wind tunnel systems controlled by a network of com-
puters and a single programmable logic controller (PLC).
The system is designed to allow multiple researchers to conduct simul-
taneous experiments in multiple wind tunnels, using wind flow generated by
a single large compressor — a 9,000,000-horsepower, 240,000,000-cubic-
foot-per-minute vacuum compressor. The system must be able to generate
air speeds of Mach 1 or greater in the experiment test section of these wind
tunnels, but also be tightly controlled so that the start-up or shutdown of
any one wind tunnel doesn’t affect the operation of any others.
Despite the sophistication of the PLC-based system that controls all the
wind tunnel interaction, its primitive operator interface (OI) provided little vis-
ibility into the compressor system’s online operation. Information about oper-
ating conditions was insufficient to allow less experienced operators — such as
doctoral candidates from around the country who use the lab for aerodynamic
research for their dissertations — to operate it as well as the best NASA operator
could.
It also was insufficient for use by FML facility engineers to analyze effi-
HLS0311_23_26_NASA.qxd 12/17/03 2:38 PM Page 23
25FALL 20 03• h o t l i n k s
A E R O D Y N A M I C R E S E A R C H
h o t l i n k s • FALL 20 0324
tem verified the state of the vanes, monitored the
surge control valve that admits atmospheric air
from outside the building, started and tested the
main and backup oil pumps, started the main drive
motor and then brought the compressor up to
speed under a low-load condition.
Once the compressor was at speed, the PLC
released the valve and the vanes under propor-
tional, integral derivative (PID) loop control and
modulated the volume of atmospheric air versus
manifold flow. The object was to maintain a fixed
pressure of 8.3 pounds per square inch absolute
(PSIA), which was ideal for providing Mach 1 air
flow, regardless of the control valve settings at the
individual test cells.
The new system was set up so that the com-
pressor always stays on-line and manifold pressure
is always available for experiments in the wind
tunnels. A General Electric 90-70 PLC MO handles
all input/output (I/O) and control functions using
a mix of I/O cards and GE Fanuc Genius blocks.
Data is passed from the PLC to the HMI computer
via a PC interface module that communicates at
153 kilobauds. Separate PID loops control the
valves and the vanes, updating data every 30 milli-
seconds.
No matter how fast the users can open or close
the valves, the compressor manifold pressure
always stays at 8.3 PSIA (+0.05) from zero to
175,000 cubic feet per minute air flow, at which
point some drift is noticeable because the compres-
sor begins to fall behind.
“The problem was that the control system was
highly reactive, so it tended to get a bit unstable,
particularly early in its development,” explains
Dave Yaste, NASA senior aerospace engineering
technician.
“We had some problems keeping the compres-
sor on set points and sometimes the IGVs and SCV
would go into oscillations and feed off each other
until the machine detected surge and shut itself
down.
“It took a lot of trial-and-error manipulation
to get it under control because we couldn’t track all
this activity other than by reading through reams of
ladder logic and trying to correlate it to columns of
historical figures. We were doing detective work,
not monitoring the process as it ran. In addition, it
was a very difficult system to learn to run simply
because we had no visibility into it.”
In 1991, a project team was created to investi-
gate the development of an advanced human-
machine interface to solve the problem. The team
included Yaste, controls engineer Peter Graube and
facility manager James Laub.
After researching more than 20 HMI possi-
bilities, the team arrived at six key features that
would be the base requirements for a successful
installation:
• Personal computer compatibility
• Statistical capability
• User-friendly interface
• Dynamic data
• Full animation
• Multi-tasking capabilities.
NASA Ames selected Wonderware’s InTouch
HMI software and NetDDE for real-time data
exchange between the InTouch HMI and the lab’s
Excel spreadsheet program via networks. This real-
time data exchange facilitates long-term trending
and statistical analysis.
The InTouch software runs on a Pentium
90 microcomputer made by EPS Technologies. The
system uses an ATI Mach 64 video accelerator,
which is configured for 1,024 x 768 resolution and
65,000 colors. The monitor is a ViewSonic 20.
Intuitive, Powerful Interface
TH E I N TO U C H S O F T WA R E H A S P R OV E N
T O B E A N I N T U I T I V E A N D P O W E R F U L
U S E R I N T E R F A C E T O T H E C O M P R E S S O R
CONTROL SYSTEM. The software displays a replica
of the compressor equipment configuration, with
appropriate operating data displayed in on-screen
windows so that operation is visible at all times,
regardless of operator familiarity with the system.
ciently any automated compressor shut-
down and determine its cause and effects.
They had to dig through extensive PLC
ladder logic to diagnose faults because
the simple OI controls that existed — dis-
crete lamps indicating on and off condi-
tions and a Panalarm light box —were
inadequate.
The problem was solved with
the addition of a PC-based, human-
machine interface (HMI) developed
using Windows-based InTouch® HMI
application development software from
Wonderware, a business unit of Invensys’
Production Management Division.
The resulting compressor and control
system now provides a sophisticated plat-
form on which NASA Ames engineers are
planning to build even more capable sys-
tems for future experiments.
The New System
TH E FML C O M P R E S S O R S Y S T E M
C O N S I S T S O F A C E N T R I F U G A L
FLOW COMPRESSOR BUILT BY ALLIS-CH A L M E R S , A 9,000-H O R S E P OW E R
S Y N C H R O N O U S M O T O R A N D A
S PE E D-I N C R E A S I N G G E A R B OX BU I LT
BY HITACHI ELECTRIC. The compres-
sor is housed in its own building adjacent
to the wind tunnel laboratory. It is con-
nected to a vacuum manifold that pro-
vides vacuum to four in-draft wind tunnel
test cells.
Variable-incidence inlet guide vanes
(IGV) open and close like venetian blinds
to provide greater or lesser air flow to each
tunnel. A 30-inch surge control valve
(SCV) opens to admit outside air, acting as
a buffer to prevent compressor damage
when the wind tunnels are turned on or
off.
The original PLC-based control sys-
The decision to move to the InTouchHMI has been an extremely profitablemove. — D A V E YA S T E , S E N I O R E N G I N E E R I N G T E C H N I C I A N
HLS0311_23_26_NASA.qxd 12/17/03 2:38 PM Page 24
One new screen is used to automate
the compressor start-up sequence, with
each step in the process causing on-screen
graphics/text blocks to change color from
yellow to green as the step is passed. In the
past, start-up was an extremely difficult
procedure because of the defense mecha-
nisms that were built in to protect the com-
pressor. Now, it starts without incident
more than 95 percent of the time. On those
occasions when any steps fail, and, there-
fore, cause a failed start sequence, problems
are instantly obvious to operators.
In addition, the Excel spreadsheet
program running in the background is
used for collecting data from about 400
system control points to build a database
for statistical analysis of trends. The lab
now allows researchers to sample any of 71
analog channels and 155 discrete points,
simultaneously animating the compressor
train and several zoom screens and storing
sample values for later analysis.
The use of InTouch HMI has allowed
the FML staff to replace a strip-chart
recorder with a four-channel, high-speed,
real-time trend program that provides an
on-screen chart view of transient data at
sample rates of up to four samples per sec-
ond. Staff can now easily compare live data
to historical data and download results to
Excel for future analysis and hard-copy
output.
As a problem-detection tool, InTouch
software allows the FML team to pop up a
screen, zoom in on any alarmed point and
show the temperatures, vibrations, pres-
sures or other variables of the components
directly related to the alarm condition.
If there is a question at any time
about operating conditions or alarm sta-
tus, the operator simply ‘walks’ through
the progressive screens to single out the
failed start step or the status of each com-
ponent. This is a far easier process than
with the old system, under which the lab
team would have to search through ladder
logic data to find the offending rung (a
sometimes cryptic troubleshooting path).
Now, the operator merely points and clicks
his way to the answer.
Adding ControlFunctionality
“T HE DECISION TO MOVE TO A
W I N D OW S - B A S E D G R A P H I -C A L HMI H A S B E E N A N E XT R E M E LY
P R O F I TA B L E M O V E , E V E N I F O N LY
MEASURED IN TERMS OF THE TROU-B L E S H O O T I N G A N D T R A I N I N G A N D
MAN-HOURS SAVED,” YASTE SAYS.“The added capabilities our new test
facility has gained, from advanced controls
and user-friendly tools, however, is
immeasurable. Even though InTouch soft-
ware is technically an HMI program, we’re
already implementing some control activi-
ties with it on one of our test stations.
“We’ve built our own stepper motor
and actuator for two tunnels, and we’re
now doing comparison studies to see
whether we should write our own DDE
using Wonderware’s DDE Server Toolkit
or use NetDDE and dynamic link libraries
to incorporate an off-the-shelf controller,”
he adds.
“Linking the screen graphics and out-
put displays to the Genius blocks and
using the logic capabilities of the InTouch
HMI will allow us to provide some very
flexible yet sophisticated control tech-
niques for customizing experiments.
“This could be very useful to candi-
dates because it would let them simply
specify the Mach number they want in the
test section of the wind tunnel, and the
InTouch HMI can calculate how far open
the choke throat should be,” he says.
“There’s a specific ratio between the
supersonic, sonic and subsonic sections by
which area ratio alone will generate a
Mach number. So merely by creating a
graphic slider on the screen, we can let the
user enter the air speed desired, and
InTouch logic will calculate and set the
tunnel parameters.”
“The success of the HMI program
and the new vistas it has opened for us has
inspired us to upgrade our PLC systems,”
Yaste notes. THE END
h o t l i n k s • FALL 20 0326
A E R O D Y N A M I C R E S E A R C H
HLS0311_23_26_NASA.qxd 12/17/03 2:38 PM Page 26