|
Source: NASA Dryden Flight Research Centers
In the 90is, NASA started to develop the Autonom Crew Return Vehicle
X-38. This vehicle is designed to bring up to 7 astronauts in an emergency
case safe back to earth.
 |
|
Fig. 1: X-38 modell
|
For this development, flight and landing test were made from increasing
dropping altitudes. From these experiments, the computer simulation can
be confirmed and refined to allow an optimization of the design. Very
important for this process is that the form of the actual model matches
the form of the object used in the simulation (CFD, Computational Fluid
Dynamics). For this reason, the actual model has to be digitized with
high accuracy and high data density during the development. From these
actual data, CAD data is generated, which is used in the simulation.
In this application note, we show how the digitizing was done and how
the data was treated to match the needs of NASA. The project was conducted
by Capture
3D as a service work for NASA.
 |
|
Fig. 2: Measuring set up
|
Digitizing was needed for the actual X-38 test vehicle, a 80% model,
approx. 33 feet long, 12 feet tall and 10 feet wide. Based on these data,
multiple evaluation tasks have to be started, e.g. "As Built Vs. As Designed",
generate an "As Built" surface model for CFD Analysis and determine the
effects of a previously "hard landing"
Why did NASA select TRITOP and ATOS over other systems
- The vehicle is a working prototype which has to be digitized as it
is. The part is too big for existing CMM's and this process is too time
consuming
- Previously work was done with a Laser Tracker, giving accurate, but
not dense data. Now NASA wanted to ensure a full vehicle definition
with ample data density on the big surfaces as well as in critical areas
- Data is needed which can support rapid surfacing from polygon mesh.
ATOS ensures direct polygonal mesh data output in different data densities,
for efficient, quick and detailed CAD generation
Special conditions for this measurement
The model has to be scanned in a working vehicle bay, during standard
work at the vehicle by the HVAC retro fit crew. No special treatment of
the object is allowed (no spray) and a full surface digitizing with high
data accuracy and high data density is needed. Bay doors are opening and
closing and ambient lighting changes will happen in addition to the object
movements. Furthermore, the underside of the vehicle has only 3 feet clearance
from the floor.
These conditions ask for a scanning system which registers the ambient
conditions, is insensitive for these influences and keep the user informed
if the conditions influence the data integrity. In addition the scanning
system has to be able to work with big areas, with the ability to be set
up to a smaller measuring are with short standoff distance do to the restricted
space under the object, to allow an efficient data acquisition.
Measurement
 |
|
| Fig. 3: Application of the markers |
Fig. 4: TRITOP measurement |
To start the measurement, coded and non-coded markers and two scale bars
are applied onto the model. In addition some "Hubbs" targets are placed
in reference bores of the model to define the coordinate system. Then
the digital images are captured which allow TRITOP to define the exact
coordinates of all markers based on the photogrammetry principle. From
this TRITOP measurement, the reference file for the detailed digitizing
is derived.
|
|
Fig. 5: TRITOP measurement
(camera and marker position) |
Fig. 6: ATOS marker position |
The digitizing of the upper and the side surfaces was made with the ATOS
HR set to a measuring volume of 800x640x640 mm.
|
|
| Fig. 7: Scan of the inner side of the wing |
Fig. 8: Scan of the outer side |
For the scanning on the underside, the scanner had to be adjusted for
a smaller measuring volume (360x280x280 mm) with a shorter stand off distance
due to the limited space under the object. The adjustment, including the
calibration of the system is finished in 15 minutes.
|
 |
Fig. 9: Individual ATOS measurement,
with ¼ of the maximal data density |
Fig. 10: dense polygonized data,
typically 6 Mio. triangles |
The captured ATOS data is automatically combined into one project. After
the digitizing, the polygonized data is automatically calculated with
the requested data density from the stored information saved during the
digitizing. Then, customized data (sections, thinned polygonized data,
feature lines) can be derived using the ATOS software and stored in different
standard data formats.
The measurement, including the data calculation and the data post treatment
in ATOS took 4 days.
Results
|
|
| Fig. 11: As Built Vs. As Designed, top view |
Fig. 12: As Built Vs. As Designed, rear view |
The color plot showing the deviations from the actual model versus the
CAD data was made in two hours. This plots show a very good fit of the
actual form to the CAD data. Only the symmetry of the wings was slightly
out of tolerance. In addition it could be shown what the "hard landing"
had not caused damaged the form of the model.
|
|
Fig. 13: Sections, with 300 mm distance,
calculated in ATOS |
Fig. 14: CAD model, made by PARAFORM
from the ATOS data |
For the CFD (Computational Fluid Dynamics), a CAD model was derived in
8 hours based on the thinned out data from ATOS. For detailed analysis,
a fine CAD model was built based on the dense ATOS data in 5 days.
The project triggered a modification on the model and was regarded from
all involved parties as very positiv.
Based on this service job, the data integrity and density, NASA decided
to buy their own ATOS XL system, to be able to integrate digitizing into
their process and be able to digitize in future also the complete and
real X-38 CRV in house.

|