{"id":333,"date":"2015-11-20T11:23:49","date_gmt":"2015-11-20T16:23:49","guid":{"rendered":"http:\/\/mrsdprojects.ri.cmu.edu\/2015teamj\/?page_id=333"},"modified":"2016-05-06T23:10:32","modified_gmt":"2016-05-07T03:10:32","slug":"mechanical-subsystem-implementation","status":"publish","type":"page","link":"https:\/\/mrsdprojects.ri.cmu.edu\/2015teamj\/system-implementation\/mechanical-subsystem-implementation\/","title":{"rendered":"Mechanical Subsystem Implementation"},"content":{"rendered":"

Final System:<\/strong><\/p>\n

\"legs\"<\/p>\n

 <\/p>\n

 <\/p>\n

12 February 2016<\/strong><\/p>\n

With the approval of the project sponsor and the course professor, the team has decided to re-scope the project to focus on the landing sequence during landing on a dynamically moving deck and remove the distance approach portion. \u00a0This change allows for the simplification to one camera, which will be a fish-eye camera, in place of the short range camera and long range camera previously used. \u00a0Additionally, using a fish-eye camera allows for fixed-mounting and the removal of the gimbal from the design. \u00a0This frees up a good amount of space to mount additional sensors (height and optical flow) beneath the DJI.<\/p>\n

New legs have been designed in CAD and fabricated using the 3D printers. \u00a0The legs have shelving slots for mounting hardware, which are much more stable than the previous hardware attachment method.<\/p>\n

\"new_legs\"<\/a><\/p>\n

The front-most shelf of the legs has space dedicated to the\u00a0height sensor, optical flow sensor, and the camera.\u00a0 The optical flow and height sensors are both downward facing with respect to the quad and the camera has several pitch settings.\u00a0 The ability to change the pitch of the camera will help with development, as the ideal angle for the fisheye camera has not yet been established.\u00a0 The pitch can be set from the camera facing directly forward to directly down, with 15degree intervals between each setting.\u00a0 It was designed for printing without support material on the 3D printer, as the use of support material can introduce significant errors in dimensions.\u00a0 The camera mount is printed in two pieces that assemble to the shelf using a press-fit.\u00a0 Side brackets are assembled to the camera, and the camera can be inserted into the desired angled slot using a press-fit.\u00a0 The small amount of compliance of the printed parts allows for snug fits and later changes to the pitch angle.<\/p>\n

\"camera_shelf\"<\/a><\/p>\n

 <\/p>\n

10 December 2015<\/strong><\/p>\n

For the Fall Validation Experiment, the short range camera was fix-mounted to the rotorcraft at a 30-degree pitch angle from the forward facing direction. \u00a0The gimbal is unusable due to the variable forces of the USB cables that attach to the cameras. \u00a0Because flying is only required within short distances of the deck, only the short range camera was needed. \u00a0The mount design for the camera is shown in the image below.<\/p>\n

\"fix_mount\"<\/a><\/p>\n

 <\/p>\n

13 November 2015<\/strong><\/p>\n

The previous design was found to be too bulky, therefore a similar design without the back panel was drafted, as shown in the picture below.\u00a0 The extra screw holes were removed, and the two remaining screw-holes will be enough to constrain the movement of the mount.\u00a0 The screw-hole locations and sizes were corrected to account for shrinking during cooling.\u00a0 The dimension between the camera side brackets was increased to account for cooling as well.<\/p>\n

\"mech4\"<\/a><\/p>\n

This model was 3D printed using the MakerBot, as shown below.\u00a0 This design allows both cameras to fit nicely between the brackets, and the screw holes are accurately placed.<\/p>\n

\"mech5\"<\/a><\/p>\n

 <\/p>\n

6 November 2015<\/strong><\/p>\n

The design of the camera mount was changed to limit the height dimension.\u00a0 Since it is a gimbal attachment, it will be important for the size to be small enough to not contact the ground beneath the DJI\u00a0when it is taking off and landing.\u00a0 Therefore, the leg length of the DJI constrains the size of the gimbal attachment. \u00a0A sandwich style gimbal mount was drafted in CAD, as shown below.\u00a0 The cameras mount in a portrait view, which also allows the received image to capture a wider range in the vertical direction.\u00a0 This is preferred, as the yaw angle can be compensated for more easily than the pitch angle by the movement of the DJI.\u00a0 The directionality of the cameras is identical, which will allow for similar transformation computations for localization.\u00a0 The IMU was removed from the design, as it was determined that the IMU data should correspond to the movement of the DJI to provide control feedback for flight control.<\/p>\n

\"mech2\"<\/a><\/p>\n

The model was 3D printed using MakerBot, as shown in the picture below.\u00a0 The dimensions of the model shrunk slightly due to cooling after printing, so screw-hole placements and sizes were not adequate.\u00a0 Also, the dimension between the side bracket shrunk and is slightly too snug to fit both cameras.\u00a0 The portion of the mount behind the gimbal is too bulky to allow for free rotation of the gimbal on the quad.<\/p>\n

\"mech3\"<\/a><\/p>\n

 <\/p>\n

28 October 2015<\/strong><\/p>\n

A CAD first draft of the camera mount was completed and is intended as an attachment for the gimbal of the DJI.\u00a0 The long range camera is oriented in the forward direction, with the short range camera angled 60 degrees down from horizontal.\u00a0 This design would allow both cameras to be mounted in a landscape orientation.\u00a0 The mount also includes space for the IMU.<\/p>\n

\"mech1\"<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"

Final System:     12 February 2016 With the approval of the project sponsor and the course professor, the team has decided to re-scope the project to focus on the landing sequence during landing on a dynamically moving deck and remove the distance approach portion. \u00a0This change allows for the simplification to one camera, which
+ Read More<\/a><\/p>\n","protected":false},"author":26,"featured_media":0,"parent":31,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"template_parts\/page-fullwidth_template.php","meta":{"footnotes":""},"class_list":["post-333","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/mrsdprojects.ri.cmu.edu\/2015teamj\/wp-json\/wp\/v2\/pages\/333","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/mrsdprojects.ri.cmu.edu\/2015teamj\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/mrsdprojects.ri.cmu.edu\/2015teamj\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/mrsdprojects.ri.cmu.edu\/2015teamj\/wp-json\/wp\/v2\/users\/26"}],"replies":[{"embeddable":true,"href":"https:\/\/mrsdprojects.ri.cmu.edu\/2015teamj\/wp-json\/wp\/v2\/comments?post=333"}],"version-history":[{"count":7,"href":"https:\/\/mrsdprojects.ri.cmu.edu\/2015teamj\/wp-json\/wp\/v2\/pages\/333\/revisions"}],"predecessor-version":[{"id":499,"href":"https:\/\/mrsdprojects.ri.cmu.edu\/2015teamj\/wp-json\/wp\/v2\/pages\/333\/revisions\/499"}],"up":[{"embeddable":true,"href":"https:\/\/mrsdprojects.ri.cmu.edu\/2015teamj\/wp-json\/wp\/v2\/pages\/31"}],"wp:attachment":[{"href":"https:\/\/mrsdprojects.ri.cmu.edu\/2015teamj\/wp-json\/wp\/v2\/media?parent=333"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}