Open Alliance
ENGLISH
Post-season
On our Github profile, we've posted our official code from Champs. Furthermore, we have released the CAD of our robot! If you have any questions towards the design, you can ask us on our social media or at robotics@porg.cz!
Saturday, 03/02/2024
We've fitted the newly decorated ramp neatly to the robot and reassembled the shooter, which had a couple of persistent flaws, ultimately fitting it to the robot in its hopefully final form. We also reassembled the drive modules in their own final form so as not to have to touch them again until the end of the season. Our other focus was the wiring.
Saturday, 02/25/2024
The consturcion of our robot is underway. We started with assembling the shooter and mounting it onto its holding structure on top of the frame. Our second order of business was to create the ramp from the to be intake to the already mounted shooter that is going to funnel the Notes through the robot.
Thursday, 02/15/2024
The 3D model is finished and we can build! Let's take a look at the most important parts of our robot and the most important dimensions.
Shooting mechanism
The shooting mechanism for the Notes is left mostly unchanged from the prototype version. We just streamlined the outlines, added some guiderails, axis of rotation and spaced the motors properly.
As you can see, the mechanism is powered by two 775pro motors geared down by 4:1 ratio. On the motors, there will be pulleys that will spin one axis of the shooter and through an array of gears, the other axle will start spinning with the same speed in the opposite direction. Thanks to using belts, we can always adjust the gearing ratio to match the top speed accordingly to our needs. Belt tensioning will be done by tensioners that we tested on our offseason robot that we built this summer.
The size of the rollers remains unchanged on 2,5 inches , therefore the gap between them is still 1,5 in and the game piece is compressed by 0,5 in. The rollers themselves are not compliant, however we might add a self-vulcanizing tape to their surface. Our testing with splicing tape weren't successful because the pieces of the tape came flying off the rollers when rotating at high speeds. The advantage of the self-vulcanizing tape is that it sticks to itself and compresses around the roller, therefore there is no glue that could come loose.
The motors are mounted to the guide plates, which guarantees that the game pieces can't get out of the mechanism as well as making sure that the game piece will always be ejected in the same way, therefore the trajectory of the shots will be more consistent. The gap between the plates is 52,8 mm and they are held in place by guiding slots in the side plates. To guarantee perfect placement, we are using custom-made gussets to hold everything together. The inner distance between side plates is 36 cm, which is the same as the size of the game piece. The aforementioned gussets will be made out of 3 mm aluminum sheet metal. The motors are held in place by custom gussets as well, exactly 2 L-brackets and 1 U-bracket.
This mechanism will be attached to the overall robot structure by two short rotational axes, which give us space in the middle of the axis of rotation of the mechanism so that we can pass the Notes through it. From last season, we know that having one long axis of rotation is not the way to go because these long axes tend to bend and therefore are very hard to replace when needed.
For this design of the shooting mechanism, there must be minimal protrusions into the space, where the game piece will be moving because we found out that it is very easy to scratch the Notes and any sharp edge within the manipulator could easily destroy the game piece. To avoid this, we will rivet the gussets and we will have to 3D print endcaps for the axis that must be secured within the inside of the manipulator.
Right now, we are manufacturing a new version of this mechanism so that we can put it on the robot itself and start testing it alongside the chassis. Once this version is complete, we will try to save as much weight on this mechanism so that we can lower our center of gravity as low as possible and so that we don't put unnecessary strain on the motor that will rotate the whole mechanism.
The rotation of the mechanism will be facilitated by a #25 chain. On the shooter side, there is a 48-tooth sprocket and on the motor, there is an 18-tooth sprocket. Considering that we use NEO 550 geared down to 64:1, one full revolution would take about 1 second, which should be enough since we will never need to make a full 360° rotation. Furthermore, we increase the force of the motor as much as possible to make sure that we will always have enough torque to rotate the mechanism.
Intaking and feeding mechanism
For simplicity, we decided that we will build the intaking and feeding mechanism into one component. Considering that we will be intaking under the frame of the robot, we don't need an articulating intake mechanism. Since we have a rotating shooter, we don't see any advantage in having an articulated feeding mechanism either. The whole mechanism is powered by a single NEO motor geared down by 4:1 ratio. Then the rotation is distributed through gears and belts to the whole mechanism. We haven't yet decided on the size of the rollers yet, however since we don't require any high-speed movement of the Notes, we will figure this on the fly. The baseplate is made from 1 mm aluminum pane but we need to find some sturdier material (very likely some polycarbonate).
The spacing of the axes is given so that the game piece is always in contact with at least one axis.
Climbing mechanism
On the sketch of the whole robot, we have a mockup of what will be our climbing mechanism. For this mechanism, we haven't yet worked on exact dimesions, however we know that it will include 50x50 mm aluminum bar and within that will be 25x25 mm aluminum bar that is supported by bearings so that it cannot get stuck inside the larger bar. We still have to design and test some latching mechanism for the chain itself.
This mechanism will be powered by a Falcon500 motor geared down by 20:1.
Chassis
We have made some changes to our chassis as well - we have fitted in some more bars so that the frame is sturdier and so that we have some space, where we can mount our structural parts. On the bottom, we milled a belly pan, where we can mount
all our electronic components. Once we have some time to spare, we will CAD out all the electronic components so that we can have predrilled holes for them once we manufacture the final version of this part. If we have enough time on the mill, we will make this part out of 4mm aluminum to lower the center of gravity of the robot even lower and so that we can tap the holes for the electronics instead of having to bolt them with nuts.
Saturday, 01/27/2024
We finally have aquired the official Notes allowing us to test our intake and shooter with them. We now also have a frame of the robot powered by swerve modules. We showed our progress to the onlooking masses as we took part in the Open doors programme of our school.
Thursday, 01/18/2024
As of right now, we still don't have the official Notes, we have built the Speaker and AMP and the we've built first prototypes of shooter and intake.
This is the current state of our intake mechanism. On the bottom, we have 1 inch rollers above which are more rollers (1in and 2,25in) which feed the Notes further into the robot.
We will very likely use a lot of 1in rollers on our robot, which we don't have in our storage. Therefore we will have to find a way to manufacture a significant amount. Right now, we are testing 3D printed rollers wrapped in 3M-???? tape which can grip the Notes well and can somewhat vulcanize when heated up. However, we have to test this a lot more within our prototypes to find out, how do these rollers compare to COTS ones. One thing we have to focus on is how well does the tape stick to the 3D print.
The intake itself seems reliable and checks all the boxes within our priority list. Next step will be CADing the final geometry of the intake so that we can test the intake within the chassis. These side plates will be made by students of ZKT (seminar of basics of mechanical engineering).
Above is the testing of our shooter in the configuration of 2 opposing 4 inch rollers with 0.5inch compression on the Note. The rollers grip the top and the bottom of the Note. In this case we ran the wheels at 3000 rpm, which theoretically shoots the Note with velocity of 10 m/s. It is important that both rollers are running same rpms to make sure that the Note exits the shooter straight.
Right now, we can't tell, how well will this work with the real Notes, however the testing of other teams shows that this shooter could be reliable. However, even though these big wheels give us a large exit velocity, they are very heavy and they don't really fit our rotating shooter concept.
So we built another prototype, which keeps the same compression on the Note but it uses 1inch rollers. A big advantage is that this shooter is smaller, lighter and can be powered by two spur gears, which make sure that the rollers spin at the same speed in opposite directions.
During the testing of this prototype, we found out that our drills aren't fast enough to achieve exit velocity high enough. Therefore, we have to wait for the official Notes and until then we can motorize the prototype. We plan on using the 775pro (redline) motor geared down by 4:1, which should give us about 5000 rpms and exit velocity of 6 m/s. This speed should allow us to shoot into Speaker from Stage, however if we wanted to shoot from closer to the middle of the field, we would have to raise the exit velocity by a different gearing or increasing the roller sizes.
Next we finalized the chassis dimensions at 68x68 cm. We need the square frame for our swerve and we wanted to minimize the frame size to fit two or three robots next to each other under the chain.
The chassis isn't built yet because of the amount of things that we were manufacturing on our CNC mill. We were manufacturing the frame, shooter and intake all at the same time. Also bad luck struck as we crashed the mill during drilling operations and we had to spend the rest of the day repairing it.
Wednesday, 01/11/2024
As usual, walking away from kickoff with a clear design in mind is impossible. Never mind the number of prototypes built in the past weekend, the US teams will find something we won't be able to notice, considering that our game pieces are now stuck in Cologne, Germany. Therefore, this recap will be short and full of rough sketches.
Firstly, we have to address our priority list, when it comes to crucial robot-design choices:
Priority list
MUST HAVE
Swerve drive chassis
Cycle time of 12 seconds (from Alliance Wall to end of field and back while handling game pieces)
Intake philosophy Touch it, Own it
Be able to pick up game pieces from the floor
Score into AMP
Score into Speaker
Be able to climb on the chain
Lift ourselves in 5 seconds
Fit 2 robots on one chain
During auto, score 1 Note and cross the line
Automatically align the robot with Notes on the ground
Automatically align the robot with AprilTags
NICE TO HAVE
Cycle time of 10 seconds
Be able to intake directly from the loading station
Be able to shoot into the speaker from both sides of the robot (back and front)
Be able to score in Trap
Fit 3 robots on the chain
Be able to lift the robot in 2 seconds
During Autonomous, score 1 Note and intake another.
Up next are our design ideas that we will iterate on during this week:
This is our no. 1 design idea, that we have worked on the most. This design works with intaking the game pieces underneath the robot. Such collected game pieces will be transported into a rotating shooting "head" of the robot, which can feed it into the AMP or shoot it into the Speaker. Other sketches can be found below.
The biggest advantage of this design is its simplicity. It involves a small number of moving pieces - the intake is static, the feeder is static, and the only moving part is the shooter. Another advantage is that the shooter can be integrated as a part of the feeder (see below)
One disadvantage of this mechanism is the rotating shooter itself. If the drivers want to change the shooter angle on the fly, the shooter would have to move with a large acceleration, which could destabilize the robot itself. One way to prevent this is to put some weight on the base of the robot and lower the weight of the shooter. If we decide to go down this path, we are considering replacing aluminum extrusions of the shooter with carbon-fiber ones.
Based on the Ri3D (Robot in 3 Days - teams of college students that challenge themselves to build a fully functional robot in just 3 days) it seems that intaking Notes from underneath the robot can be troublesome, therefore we have to build our chassis ASAP and finish our intake prototype to be able to determine via testing if we can make this design work or if we have to design an over-the-bumper intake.
Another issue is that this design uses a fairly long feeder mechanism (over 50 cm ~ 20"), which will have to have the feeding elements close to each other so that the Note doesn't get stuck in between these elements. If we were using rollers, we would soon run out of compliant wheels, therefore, we are experimenting with putting some rubber on 3D-printed rollers, which would replace the compliant wheels.
In this sketch, there is a rough calculation of the minimal needed feeder length and the idea of a rotating shooter (in yellow) is shown.
Osa rotace = axis of rotation
The other way how to be able to score game pieces is to hang the shooter onto 2 independent elevators that would be able to precisely set any angle that we desire. This idea has to be further explored so that we can draw some conclusions and eventually build. Below is a very rough sketch of this idea:
The biggest advantage of this design is that we could mount the climber structure onto one of the elevators.
For the climber part, we plan to use a structure based on Andymark's Climber in a Box (image below). We designed a very similar structure 2 years ago when we managed to create a working climber without our CNC mill. With more years of experience, we believe that we will be able to implement this without any issues.
If we use this structure, we will have 2 climber arms, where each would have its winch powered by a Falcon 500 motor geared down by a 1:20 gearbox. Thanks to this, one arm should be able to lift the robot in 2 seconds. With such quick lifting, we need to find a way to quickly latch the robot onto the chain so that we can score game pieces as long as possible.
If we use the idea with 2 elevators, we would mount the latching structure to the lifting structure and use a climber of our design.
For latching the robot to the chain, we are testing the use of a "comb" consisting of bolts, that can latch into the links of the chain and prevent the robot from sliding down the chain. In the next post, we will link photos of our prototype models and a model that reflects the real chain.
Illustrative photo, source
In terms of the drivetrain, we plan to use our SDS MK2 modules, which we managed to revive for this season. We planned to manufacture our modules based on the Swerve X from West Coast Products, but this got very delayed, and right now, we want to focus on manufacturing the prototypes of the robot.
Considering our plan to intake game pieces from underneath the robot, we feel that the need for inverted modules is eliminated because we don't need to use the space above our modules for anything specific, such as an intake structure.
Both the CAD model and the physical chassis should be built this week, so we will post photos of both in the next post.