A Soft 3D-Printed Robotic Hand Actuated by Coiled SMA

Round 1

Reviewer 1 Report

In this paper, a 3D-printed soft robotic hand actuated by coiled shape memory alloy is proposed. Compared with the traditional robotic hand, this hand has several advantages; low-cost, simple assembly, ease of manufacturing, and lightweight. The experiments show that this hand can grasp several daily objects, and thus it seems to be useful. However, this paper has several unclear and insufficient points as follows.

1. From Table 2, the initial length of each muscle differs. That is, the individual difference exists and affects the muscle performance investigated in section 4.1. The reviewer recommends that the authors discuss and evaluate this point in section 4.1.

2. In section 4.2 and section 4.3, it is unclear why the test currents were not unified. That is, the currents were set to 0.4, 0.5, and 0.7A in section 4.2, while those were set to 0.4, 0.5, and 0.7A in section 4.3.

3. In section 4.4, why did not the problem of the index and little fingers described in section 4.3 happen? The authors should add a description of this point.

4. Although the causes why the index and little fingers are difficult to bend were discussed in section 4.6, those should be discussed in section 4.3 at first.

Those followings are minor comments.

5. Please edit Figures 4, 5, 6, 7, 10, 11, 12, and 13. In those figures, the places of captions ((a), (b), etc.) are strange. Furthermore, Figure 8 does not exist in this paper.

6. The section number of the conclusion does not exist.

7. In Figure 12, unifying the color bar will make it easy to compare each stress result.

Author Response

Reviewer 1

Comments and Suggestions for Authors

In this paper, a 3D-printed soft robotic hand actuated by coiled shape memory alloy is proposed. Compared with the traditional robotic hand, this hand has several advantages; low-cost, simple assembly, ease of manufacturing, and lightweight. The experiments show that this hand can grasp several daily objects, and thus it seems to be useful. However, this paper has several unclear and insufficient points as follows.

1. From Table 2, the initial length of each muscle differs. That is, the individual difference exists and affects the muscle performance investigated in section 4.1. The reviewer recommends that the authors discuss and evaluate this point in section 4.1.

Response 1: The authors would like to thank the reviewer for the input. These differences are due to variations in the manufacturing process outlined in section 3.2, in which muscles are produced in 50 mm lengths, and then divided in two, to form muscles of 25 mm length each. When cutting the muscles there may be slight deviations, as noted in Table 2.  However, due to the pre-tensioning method described in section 3.3, the differences of ±2-3 mm is effectively negated by the much larger pretensioned length of 100 mm. Each muscle attempts to recover to its original length via shape memory effect, but it is limited to a maximum contraction of approximately 50 mm due to the geometry of the finger and placement of the muscles.  This description has been added to the text.

1. In section 4.2 and section 4.3, it is unclear why the test currents were not unified. That is, the currents were set to 0.4, 0.5, and 0.7A in section 4.2, while those were set to 0.4, 0.5, and 0.7A in section 4.3.

Response 2: Safe and effective actuation is a combined effect of heating time and actuation current magnitude, as heating for a long time at higher current will damage the muscles, whereas heating at higher current for short period will not damage the muscles as such.   In section 4.2, it is noted that “0.7 A was not utilized in the preliminary tests to avoid damaging the muscles due to the long constant heating times of 5 and 10 seconds.” However, in Section 4.3, the heating time of the flexor is only 2 seconds at 0.7 A, and thus does not risk damaging the muscle. The difference in heating times is due to the sub-optimal heating times exhibited in Figure 8b and c, in which the finger is not actuating for long periods of time between cycles (shown in the constant regions of 0 degree actuation in the graphs) while the muscles are still being heated. When the heating time is only 2 seconds, 0.3 A is insufficient to demonstrate observable actuation, so this was increased to 0.4 A. Changing the heating cycles also allowed for faster actuation cycles, later shown in Section 4.6 (frequency tests), where actuation cycle times were decreased to 8 seconds, as opposed to the 15 second cycles shown in section 4.2. Parts of this explanation has been added to the text as well.

1. In section 4.4, why did not the problem of the index and little fingers described in section 4.3 happen? The authors should add a description of this point.

Response 3: In section 4.3, the fingers are not grasping any objects and thus close as far as the muscles allow. In section 4.4, the objects being grasped limit the amount of distance the fingers can close. In addition, section 4.3 notes that the first cycle has non-problematic actuation, whereas later cycles experience issues. Contrarily, section 4.4 is a non-cyclic test, and thus does not encounter any problems due to each object being grasped in their own “first cycle,” with no further actuation beyond the first. This explanation has been appended to section 4.4 for clarity.

1. Although the causes why the index and little fingers are difficult to bend were discussed in section 4.6, those should be discussed in section 4.3 at first.

Response 4: The authors thank the reviewer for their input, and the change has been implemented as suggested.

Those followings are minor comments.

1. Please edit Figures 4, 5, 6, 7, 10, 11, 12, and 13. In those figures, the places of captions ((a), (b), etc.) are strange. Furthermore, Figure 8 does not exist in this paper.

Response 5: The authors believe this was due to an error when the original uploaded manuscript was converted to the MDPI template. Figure 8 appears in the second uploaded manuscript that was submitted in conformance with the MDPI formatting guidelines/template. It has been checked to be in the revised version.

1. The section number of the conclusion does not exist.

Response 6: Similar to above, the section number of the conclusion was likely accidentally truncated when the original manuscript was converted to the MDPI template.  It has been checked to be in the revised version.

1. In Figure 12, unifying the color bar will make it easy to compare each stress result.

Response 7: The authors thank the reviewer for the input, and the change has been implemented as suggested. The displacement result has been unified as well.

Author Response File: Author Response.pdf

Reviewer 2 Report

This article is the first time to use 3D printing technology to print TPU material to construct the shape of manipulator, which is a relatively new point. In addition, the product adopts the overall lightweight structure, which is easy to assemble and low in cost, which is also the advantage of the product.The following are some suggestions for the author. 

1. There is something wrong with the layout of the article. Some titles and texts are separated on two pages, and tables and pictures are randomly placed.
2. Figure 2 The muscle behind the Antagonistic Coiled SMA Extensor is not completely written.
3. C in Figure 4 is not marked.
4. The labels of Figure 5, Figure 6, Figure 7 and Figure 10 do not correspond to the figures. Letters do not correspond to pictures.
5. Picture 8 is missing in the article, please add it.
6. The effective figures after decimal point are inconsistent.
7. It is suggested to increase the simulation of applying force perpendicular to the manipulator surface to illustrate the bearing capacity of the manipulator.

Author Response

Reviewer 2

Comments and Suggestions for Authors

This article is the first time to use 3D printing technology to print TPU material to construct the shape of manipulator, which is a relatively new point. In addition, the product adopts the overall lightweight structure, which is easy to assemble and low in cost, which is also the advantage of the product. The following are some suggestions for the author. 

1. There is something wrong with the layout of the article. Some titles and texts are separated on two pages, and tables and pictures are randomly placed.

Response 1: The authors believe this was due to an error when the original uploaded manuscript was converted to the MDPI template, as the formatting was specifically checked in the second uploaded manuscript that was submitted in conformance with the MDPI formatting guidelines/template. Please see the revised PDF version in case the revised Word document has persisting formatting problems.

1. Figure 2 The muscle behind the Antagonistic Coiled SMA Extensor is not completely written.

Response 2: The figure may have been cut off accidentally by the formatting issues mentioned above. The figure has been checked to be resized; please see the revised PDF version.

1. C in Figure 4 is not marked.

Response 3: This also appears to be a formatting issue. Please see the revised PDF version.

1. The labels of Figure 5, Figure 6, Figure 7 and Figure 10 do not correspond to the figures. Letters do not correspond to pictures.

Response 4: This also appears to be a formatting issue. The captions have been checked for correctness again; please see the revised PDF version.

1. Picture 8 is missing in the article, please add it.
   
   

Response 5: Similar to Response 1, Figure 8 appears in the second uploaded manuscript that was submitted in conformance with the MDPI formatting guidelines/template. It has been checked to be in the revised version.

1. The effective figures after decimal point are inconsistent.
   
   

Response 6: Significant figures were recorded to each instrument’s resolution. Calculations were rounded to two significant digits. We have edited decimal point as much as possible.

1. It is suggested to increase the simulation of applying force perpendicular to the manipulator surface to illustrate the bearing capacity of the manipulator.
   
   

Response 7: The authors thank the reviewer for their input, and the suggestion has been implemented, and a new figure (Fig.13) is added. An explanation has been added to the text as well.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Based on all reviewer’s comments, the paper is well-revised.
The reviewer thinks that it is suitable for publication.

Reviewer 2 Report

The topic selection of this article is relatively new, the work is more detailed, and the theoretical support is sufficient.The simulation and design problems have been modified.It is suggested that the research be accept in present form.