The Shack-Hartmann Demonstration
Introduction About AO Lens Vision Ray Tracing Contact Us
The Shack-Hartmann Sensor in an AO System
The Shack-Hartmann sensor plays a very important role in an adaptive optics system. The Shack-Hartmann sensor detects the optical distortions the Earth's atmosphere causes to images of objects in space, and enables the distortions to be measured and corrected (more on the SH and AO).
Demonstrating the Shack-Hartmann Sensor (generation 1)
Our demonstration shows the image formation process of the Shack Hartmann sensor. In the demonstration a light source (in the shape of a star's diffraction pattern) is projected at an array of 16 lenses. A target is placed at the focal length of the lenses, with the ideal centers for each image marked. When there is nothing between the light source and the lenses, the centers of all 16 images line up with their ideal centers. To demonstrate the atmosphere, a clear tray filled with mineral oil is placed between the lenses and the light source. Then, surface waves are created in the tray. As a result, the images on the target can be clearly seen moving in different directions at different speeds.
The schematic setup is as follows:
Figure 4: Schematic set-up for our demonstration.
Constructing the Demonstration
|
Light Source The light source was made from a wooden box with an open face. The box was 12'x12'x13'. The inside of the box was covered with aluminum foil to reflect as much light as possible towards the open side. Two 100-watt light bulbs were connected inside the box. On the face of the box there was a piece of ground glass to diffuse the light, and in front of that a cutout of the desired image was placed. |
Figure 5: The Light Source with its Final Face. |
|
Figure 6: A Diffraction Pattern from a Circular Aperture.13 |
We originally used an arbitrary cutout for the face of the box, and used that to determine at what distance the image could be clearly seen at a decent size. The optimal distance for our purposes was 25'. Eventually, we designed a new front for the box. The front (the cutout) was made from black cardboard. The front was in the shape of a circular aperture diffraction pattern, in order to show what the image of a star would look like through a telescope. To design the front, we first measured the radius of the rings on an actual diffraction pattern. We then scaled them up so that at the given distance (25') from the light source to the lenses, the pattern would fit inside of the 0.2" x 0.2" box at the center of the target (see Target). |
We designed the cover on AutoCAD LT. Then, we cutout the rings and glued them to wax paper so that they would keep their shape and not interfere with the light exiting the box between the rings. We used Velcro to hold the front to the box so that it could be adjusted as needed.
Lens Holder
| The lenses we used had a diameter of 1.8" and a focal length of 1.5'. In order to design a holder for the lenses, we first drilled some approximately sized holes 0.5", 1", and 1.5" apart. We then put lenses in the holes and set up the light source 25 feet away. We determined that for the given distances, the images looked best when they were 0.5 inches apart. Next, a grid was designed on AutoCAD LT. It was 10" square with the circles in a square array. A 1" spacing was left on each side and 0.5" between the circles. |
Figure 7: The Layout of the Lens Holder. |
|
Figure 8: The Finished Lens Holder. |
A drill press was used to drill appropriately sized holes, .5 inch apart, in a piece of 10"x10"x3/4" wood. After the holes were drilled they were sanded down to exact size, since there was a slight variation in the sizes of the lenses. The lenses were held in place with cable ties on each side. We wanted to be able to move the lenses slightly if necessary, to account for slight variations in their focal lengths. The cable ties worked very well for this as long as each one was cut to the exact size specific to its hole. |
If the light source was at infinity, the incoming light would be parallel and the center of the lenses would line up exactly with the ideal centers of the images. However, since our light was a finite distance away this had to be accounted for when designing the target. It is important for the target to be very precise so that the images will be exactly on center when the light is coming in parallel.
| We accounted for using a light source
at a finite distance by finding the deviation of each image from its ideal
center. We then created a modified target. The process is shown below.
L is the distance from center of the lens holder to the center of the light source. d is the distance from the center of the lens holder to the ideal center. x is the distance between the center of the lens holder and the ideal center. y is the distance the actual image is off center. |
Figure 9: Finding the Ideal Centers |
By similar triangles y / x = d/L, therefore y = xd / L.
The radial deviation of the image can be found by adding y + d. Then, using AutoCAD LT, the ideal template was drawn and then radial lines were drawn through the center of each circle from the midpoint of the lens holder. The y + d distance was used as the radius of three new circles.
|
Figure 10: The Layout for the Target. The new center lines were at the intersection of the new circles and the radial lines. At each intersection a target symbol was placed with marks every tenth of an inch for one inch. |
Figure 11: The Finished Target. Once the target was design, it was printed out , laminated, and attached to a 10" x 10" x 3/4" piece of wood. |
Atmosphere
In the demonstration it was essential that we were able to simulate the distortions caused to the wavefronts by the Earth's atmosphere. Our first idea for a distortion was to make a piece of warped plastic that would rotate in front of the lenses. This worked, but it also caused the images to go in and out of focus. We then tried using a hot plate and a hair dryer to distort the light, thinking that the heat would cause the desired result. However, the heat did not affect the images. After consideration, we determined that the heat moved the air molecules over too short of a distance relative to one another, and we needed longer surface waves to simulate the Earth's atmosphere.
| In order to create these waves we made a shallow tank with a clear bottom. This was filled with an inch of mineral oil. When the tank is rotated between the light source and the array of lenses, the images on the target move very noticeably in different directions. Eventually we may use a motor to rotate the tank, but for now this is done by hand. |
Figure 12: The Atmosphere Simulator. |
The Final Setup
Figure 13: The Shack-Hartmann Demonstration at its finest!
Demonstrating the Shack-Hartmann Sensor (generation 2)
The second generation Shack-Hartmann Sensor more clearly demonstrates the function of the Shack-Hartmann sensor. This demonstration shows the viewer what the incoming image of a star looks like to astronomers, and how the lens array reduces the image into it's component parts--from which useful information can be gained. When viewers look at one screen they see the incoming image, which is blurry and distorted. When viewers look at the other screen they see sixteen pinpoint images, all moving in different directions.
| In the second generation sensor, a light source shines light onto the
mirror at the top, as it did in the generation 1 demonstration, and the
light is reflected though a moving tray of liquid that simulates the
Earth's atmosphere (not shown in the picture). Then, the light
travels down through a lens and focuses onto a beam splitter. Half of the light
is projected onto a screen beneath the beam splitter and shows the
incoming image of a star. The other half of the light is collimated and then travels though
the array of sixteen lenses. Each of these images are projected onto
a screen.
Figure 14: Generation 2, the SH demonstration. This set-up allows users to see first-hand the effect of the Earth's atmosphere on light, and therefore why adaptive optics systems are so valuable. Through the contrast between the blurry, distorted image and the sixteen moving pinpoint images, viewers can see how the Shack-Hartmann array provides astronomers with quantifiable data that can be used to reconstruct the incoming wavefront. |
Figure 15: The Finished Generation 2.
|
The whole adaptive optics process can seem very abstract. Demonstrations like this one allow observers to see an enlarged version of what happens on a microscopic level inside an adaptive optics system in a very concrete way. This demonstration is simple enough for the everyday nonscientist to understand, yet it clearly illustrates the role of the Shack-Hartmann detector in an adaptive optics system.