This week, we were introduced to basic methods of fastening and attaching that will help us produce 3 dimensional contraptions with multiple parts. These methods included machinery as well, and we learned how to properly and effectively operate them. We explored hinges, heat staking, pegs and notches, and rods and bushings.
Hinges: We constructed a hinge made of two Delrin parts joined by a piano wire. By lining up the components and drilling a hole vertically throughout the pieces, we were able to pass piano wire through the holes and create a hinge. This method is a good way of fastening components that aim to have movement relative to each other, however due to this freedom of movement, it is not a good method to use when desiring a stationary relationship between the parts. Hinges with piano wire would be ideal in situations where a loose rotating object is desired.
Heat staking: We fitted a piece of Delrin with a tab into another piece of Delrin with a cutout. The pieces intersected perpendicularly. We essentially melted the Delrin tab protruding from the cutout so that due to its new shape and attachment to the piece with the cutout, the piece with the tab cannot be separated. This is a good method for improving the strength of the connection between two pieces of Delrin by fusing them together. It does not supply for freedom of movement, however. This method would be especially helpful when fastening bases and supports together, as it will increase the structural integrity and will not come apart. We also found that heat staking a Delrin rod to a sheet of Delrin is also possible, but requires adjustments to the heat-staking process due to the quality of the Delrin being glass-reinforced.
Tabs and notches: We explored how a piece of Delrin with a tab can be inserted into another with a cutout. Given the width of the Delrin sheet and the width of the notch, the two can be fitted together with varying freedom of movement. For a given Delrin sheet width of x, a cutout width of a value ~0.01mm + x creates a snug fit. However, if the width is x or less than x, the tab cannot fit into the notch. If the width is a larger degree bigger than x, then the tab can enter the notch but will have relative movement, as a larger width gives rise to a larger tolerance and a looser fit. This method is helpful when one needs a tight fit between two components that have widths that are very close in value. It is not very helpful when the width of the notch is larger than the width of the sheet, as it would make the connecting juncture looser than what may be needed for a strong structure. This method would be great for joining components that need to be disassembled for transport or other reasons.
Rods and bushings: We found this method to be very similar to the tabs and notches method. The rod is inserted into a bushing, whose relative diameter to the rod's determines the tightness of the fit. If the diameter of the bushing is slightly larger than that of the rod, the bushing is tight. As the diameter increases, then the bushing will be looser on the rod until it stops touching it on all sides altogether. This method is beneficial when one needs to insert a cylindrical component into another surface with a tight fit. Like the tabs and notches, this method fails when the bushing's diameter exceeds that of slightly greater than the rod's and the rod can either slip out or be rendered useless. This method is great to consider when designing a simple axle.
The search to find loose vs. tight led us to measure the tolerances for the difference between the width of the inserted part and the opening.
We began with the purpose of having different degrees of freedom. For both tabs and rods, looser interactions with the surrounding part enabled slight movement and tighter interactions restricted movement to be near stationary. Tight bushings on the rod are needed to hold the rod in one place relative to the bushing, and to keep it stable. Looser bushings allow for a larger range of movement, which may be useful in designs which utilize internal lateral and vertical movement.
So what is the difference that determines a "loose" from a "tight" fit? It is the relationship between the material width of the piece inserted and the width of the opening. With digital calipers, we measured the tolerances for loose and tight fits for the tabs and notches as well as the rods and bushings.
Using the tab and notch table number 1, we measured the width (vertical width) of the notches and compared it to the material width of the tab. We found the tab material width to be 3.13 mm. The first row of notches proved to be very snug fits by physical testing. When measuring with the calipers, the average reading was 3.15 mm as the width of the notch. The second row yielded a much looser fit (the tab could be placed in and taken out with ease), with an average reading of 3.45 mm. The third, fourth, and fifth rows all showed a progressive increase in tolerance, with readings of 3.66, 3.97, and 4.22mm, respectively. As the tolerance is the difference of these readings from the material width of the tab, the tolerance grew as the notch width grew, and the junctures became progressively looser.
The rod and bushing example yielded similar results. The diameter of the rod was measured by the calipers to be 6.33mm. The bushing provided that was the tightest had an inner diameter of 6.46mm, and the loose bushing had a inner diameter of 6.66mm. This shows that 0.13mm and 0.33mm of a difference in diameter is a range of the differences of a "tight" vs "loose" bushing.
We then investigated the exactness of computer modeling as it is translated through the laser cutter to create an actual three dimensional piece. If a piece is modeled to be 0.1 cm long, will it actually be exactly 0.1 cm long when it is printed? Examining table 2 shed further light on this exploration.
Row one claimed to be 0.135" long in Solidworks. The three notches, from left to right, showed vertical widths of 0.145, 0.1435, and 0.1435". The greatest discrepancy is between 0.135" and 0.145". This is a difference of 0.01".
Row two at 0.125" showed values of 0.135, 0.1340, 0.1340". The maximum difference is 0.01".
Row three of 0.115" had values measuring 0.1285, 0.1185, and 0.1190". The maximum difference is 0.014".
It appears that on average, the printed length is 0.01", or 0.254mm, greater than what is modeled on the computer. This tolerance difference may be a result of the nature of the laser cutter. Although the cutter is set to a hairline width, the laser cannot move fast enough and cut through all the material instantaneously to recreate the actual dimensions in Solidworks. To cut entirely through the Delrin, the laser cutter speed is slowed. This may be the reason why the dimensions of the holes are bigger when printed than on the computer, as a slower speed melts more Delrin at the cut site, which could in turn explain the loss of 0.01". I would predict that the material thickness does impact the tolerance. A thicker sheet of Delrin would require a slower laser speed in order to cut through all the material in one pass. As the laser slows, more of the Delrin at the cut site would be impacted, and the difference in tolerance would be greater than that of a thinner sheet of the material. This brings up the issue of predicting and testing tolerances when material sheets of different widths are used. When using a thicker sheet of material, one should anticipate and measure a larger tolerance than that of a thinner sheet.
No comments:
Post a Comment