README.md 16 KB
Newer Older
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
1
# COVID-19 Swab Testing
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
2
3
# ***MENU***

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
4
 Introduction and Allowable values [^1]
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
5
6
7



Alfonso Parra Rubio's avatar
v2    
Alfonso Parra Rubio committed
8

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
9
In this page all design proposed as swabs are going to be tested. Thanks to Martin Culpepper we could receive 4 NP Swabs and those are being tested to serve as a reference for future models done.
Alfonso Parra Rubio's avatar
v2    
Alfonso Parra Rubio committed
10

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
11
The tests to determine the viability of any design will be:
Alfonso Parra Rubio's avatar
v2    
Alfonso Parra Rubio committed
12

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
13
  - **Bending Test**
Alfonso Parra Rubio's avatar
v2    
Alfonso Parra Rubio committed
14

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
15
  Consisting on bending the swab 35 degrees while measuring the vertical component of the reaction force.
Alfonso Parra Rubio's avatar
v2    
Alfonso Parra Rubio committed
16
17


Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
18
  - **Tip Compression Test**
Alfonso Parra Rubio's avatar
v2    
Alfonso Parra Rubio committed
19

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
20
  Compression test to determine the resistant force to be compressed of the hairy tip of the swabs when compressed until the nominal value of the hard stick that holds it. It is wanted to compress only the tip, not the core.
Alfonso Parra Rubio's avatar
v2    
Alfonso Parra Rubio committed
21
22


Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
23
24
25
26
27
28
29
30
31
32
  - **Final Tension Test**

  Destructive tension test that sets the failure point when submitted to a tension force.

  - **Axial Torsion Fatigue Test**

  Static load test that determines the value of the angle in which a constant tension start decreasing when rotating the swab.


**IMPORTANT!** Two more test could be designed. Those one will be related with the amount of particles collected in the tip.
Alfonso Parra Rubio's avatar
v2    
Alfonso Parra Rubio committed
33
34


Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
35
36
[^1]: # **Allowable values**. Testing official NP Swabs at CBA Instron

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
37
38
<img src="images/OffNPS.jpeg" alt="plot1" width="900"/>

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
39
40
41
42
43
In this section, results of all experiments will be posted and can be used to compare with any value obtained by any proposed design. With difficulty, we have been able to receive 4 samples. That's why it was not possible to make multiple times some tests to establish a mean value and a deviation.

## **Bending Test**

Tooling was designed to this purpose.
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
44
45
<img src="images/mount.jpeg" alt="plot1" width="1000"/>

Alfonso Parra Rubio's avatar
v2    
Alfonso Parra Rubio committed
46
47
48

![gif](images/test.gif)

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
49
50
51
52
53
54
55
56
The movable part of the Instron will compress the swab in a cantilever, making it to bend until the upper part of the tooling touches the lower part of the tooling. We will be able to see in the data when the step in the force happens, being able to see the previous instant which will correspond with the force tha the swab makes under a 35 degrees bending.

The official NP Swab mounted in the tooling looks like the following:

![NP Bending](images/NPBending.jpeg)

Results obtained for this test are:

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
57
<img src="images/npbend.png" alt="plot1" width="500"/>
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
58
59
60
61
62
63
64
65
66
67


Two clear region can be seen. The maximum load that this first linear slope reaches is 0.18N. Up to that it start growing linear but with a much higher slope until 1.2N value is reached.


## **Tip Compression Test**

This test consist on compressing the hairy tip of the swab. As we don't want to compress the main stick that holds the tip, the value of the thickness will be our criteria to stop the test.

The lay-up shows as follows here:
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
68
69
<img src="images/compressedNP.jpeg" alt="plot1" width="900"/>

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
70
71
72
73


The results are the following:

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
74
75
<img src="images/npcomptest.png" alt="plot1" width="500"/>

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
76
77
78
79
80
81
82
83
84
85
86


Exponential grow can be seen reaching a maximum value of 2.5N when reached the nominal thickness of the stick. Similar behavior has been seen with cotton swabs in this same test.


## **Final Tension Test**

An NP Swab has been submitted to an ultimate tensile stiffness until failure. The geometry of the NP swab and the material will determine the behavior of this test. As it is monolithic and solid we will be able to see a classic elastic deformation reaching to a plastic deformation. The value of this second slope will be determined by the material.

The lay-out of this is the following:

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
87
<img src="images/NPTension.jpeg" alt="plot1" width="900"/>
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
88
89
90
91


And results are the following

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
92
93
<img src="images/nptensiontest.png" alt="plot1" width="500"/>

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
94
95
96
97
98
99
100
101
102
103

The swab has been able to deform 16mm. The marks of the clamping system were statics so the test is valid. Also, the deformation compared with another partner (after the natural shrink after breaking) is huge!

![comp](images/NPelongated.jpeg)


## **Axial Torsion Fatigue Test**

This test has been tricky due to our machine is only available to make linear displacements. I have designed and manufactured tooling to be used in tension but able to have a degree of freedom, the rotation along this same axis.

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
104
105
<img src="images/NPTension1.jpeg" alt="plot1" width="900"/>

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
106
107
108
109
110
111
112
113
114

The test is described as following: I have pre-tensioned the swabs into a 8N force. I have started a tension test in the instron with a feed of 0.1mm/min (almost nothing). I have been rotating and counting the tooling until the tension force dropped suddenly in the test. when this happens, we have broke internally the swab by this fatigue torsion caused.

Tooling designed for the test is here:

![tooling](images/GigsToolrotate.jpeg)

![assyrot](images/assyrot.jpeg)

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
115
![rotatingGig](images/rotatingGiG.gif)
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
116
117
118
119
120
121
122
123
124
125
126
127
128
129


![gif](images/axialtest.gif)

The result for this test is, the tension has dropped up to 2N suddenly when I have passed a rotation of **710º**.

|                   | **NP Swab** |
|:-----------------:|:-----------:|
| **Angle Reached** |     710º    |





Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
130
# **First Candidate**. OPT Industries Swab by Jifei.
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
131
132
133
134
135
136
137
138
139
140
141
142

Jifei provide us 16 models to be tested.

![samples](images/samples.jpeg)
![samples2](images/samples2.jpeg)
![samples](images/jifeisamples.jpeg)

I have developed to them all tests described bellow. Here you can find the results.

### **Bending Test**

Results for all elements are the following. Raw data can get download on the files folder.
Alfonso Parra Rubio's avatar
v2    
Alfonso Parra Rubio committed
143
144
145

#### Sample 1

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
146
147
<img src="images/plot1.png" alt="plot1" width="500"/>

Alfonso Parra Rubio's avatar
v2    
Alfonso Parra Rubio committed
148
149

#### Sample 2
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
150
<img src="images/plot2.png" alt="plot1" width="500"/>
Alfonso Parra Rubio's avatar
v2    
Alfonso Parra Rubio committed
151
152

#### Sample 3
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
153
<img src="images/plot3.png" alt="plot1" width="500"/>
Alfonso Parra Rubio's avatar
v2    
Alfonso Parra Rubio committed
154
155

#### Sample 4
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
156
<img src="images/plot4.png" alt="plot1" width="500"/>
Alfonso Parra Rubio's avatar
v2    
Alfonso Parra Rubio committed
157
158

#### Sample 5
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
159
<img src="images/plot5.png" alt="plot1" width="500"/>
Alfonso Parra Rubio's avatar
v2    
Alfonso Parra Rubio committed
160

Alfonso Parra Rubio's avatar
v3    
Alfonso Parra Rubio committed
161
162

#### Sample 6
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
163
<img src="images/plot6.png" alt="plot1" width="500"/>
Alfonso Parra Rubio's avatar
v3    
Alfonso Parra Rubio committed
164
165
166


#### Sample 7
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
167
<img src="images/plot7.png" alt="plot1" width="500"/>
Alfonso Parra Rubio's avatar
v3    
Alfonso Parra Rubio committed
168
169

#### Sample 8
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
170
<img src="images/plot8.png" alt="plot1" width="500"/>
Alfonso Parra Rubio's avatar
v3    
Alfonso Parra Rubio committed
171
172

#### Sample 9
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
173
<img src="images/plot9.png" alt="plot1" width="500"/>
Alfonso Parra Rubio's avatar
v3    
Alfonso Parra Rubio committed
174
175

#### Sample 10
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
176
<img src="images/plot10.png" alt="plot1" width="500"/>
Alfonso Parra Rubio's avatar
v3    
Alfonso Parra Rubio committed
177
178


Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
A plot comparison of all results vs. NP swab will be done. For now, this table will serve us as a first comparison.

|                               | **NP Swab** | **OCT Industries** |
|:-----------------------------:|:-----------:|--------------------|
| resultant force at 35º (mean) |     1.2N    |        0.27N       |

Despite the difference in the ultimate value, the initial slope of bending is so similar and well worth to compare

|                                | **NP Swab** | **OCT Industries** |
|:------------------------------:|:-----------:|--------------------|
| resultant force at 16mm (mean) |    0.18N    |        0.113N        |




Alfonso Parra Rubio's avatar
v3    
Alfonso Parra Rubio committed
194
195
After watch 10 test of this elements, it can be seen that there is a linear part of the test that represent the stiffness of the beam but suddenly, in all samples, the slope changes and it represent the compression of the tip of the swab.

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
196
197
The elements 6,7,8,9, and 10 (different curing parameters) can be seen that the resin cycle has affect slightly to an increment of stiffness. From a 0.125 N mean value of the 5 first elements, the next 5 elements shows a mean value before the commented slope of 0,164 N.

Alfonso Parra Rubio's avatar
v3    
Alfonso Parra Rubio committed
198

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
199
### **Tip Compression Test**
Alfonso Parra Rubio's avatar
v3    
Alfonso Parra Rubio committed
200
201


Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
202
Swab tips were provided as well as cotton tips to compare this same test as described bellow. Jifei design has void tips, making the deformation more linear than Official NP swabs and cotton swabs.  
Alfonso Parra Rubio's avatar
v3    
Alfonso Parra Rubio committed
203
204
205
206
207
208
209
210
211
212
213
214
215
216

The samples provided are:

![tipSwabs](images/comp.jpeg)


The two OCT samples were tested as well as the 4 cotton tips of the cot1 and cot2 elements.

The lay up of this test is the following:

![test2](images/test2.gif)

Results for the sample done by OCT Industries are:

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
217
<img src="images/oct1.png" alt="plot1" width="500"/>
Alfonso Parra Rubio's avatar
v3    
Alfonso Parra Rubio committed
218

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
219
<img src="images/oct2.png" alt="plot1" width="500"/>
Alfonso Parra Rubio's avatar
v3    
Alfonso Parra Rubio committed
220
221
222
223


Can be seen that the compression is very linear and in both elements, under a compresion of 1.3mm the value is 8N.

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
224
For the cotton tips, results are interesting.
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
225
<img src="images/oct1.png" alt="plot1" width="500"/>
Alfonso Parra Rubio's avatar
v3    
Alfonso Parra Rubio committed
226

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
227
<img src="images/cot2.png" alt="plot1" width="500"/>
Alfonso Parra Rubio's avatar
v3    
Alfonso Parra Rubio committed
228

Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
229
230
231
<img src="images/cot3.png" alt="plot1" width="500"/>

<img src="images/cot4.png" alt="plot1" width="500"/>
Alfonso Parra Rubio's avatar
v3    
Alfonso Parra Rubio committed
232
233
234
235
236
237


As cotton cant ensure a fixed geometry, the results has a biggest deviation. Also, as we are accumulating mass when compressing, the grow of the force seems more likely as an exponential. OCT design doesn't behave the same due to its internally hollow.


Files and python script can be found inside the project folder.
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252

Comparison with official results are here:

|                               | **NP Swab** | **OCT Industries** | **Cotton** |
|:-----------------------------:|:-----------:|--------------------|------------|
| resultant force at ult (mean) |      7N     |        9.1N        | 11.2N      |


### **Tension Test**

Swabs number 11 12 and 13 were mounted in the tensile clamps of the instron to develop the test.

![tension](images/JiffeiTension.jpeg)

Results are the following:
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
253
254
255
256
257
258
<img src="images/tension11.png" alt="plot1" width="500"/>
<img src="images/tension12.png" alt="plot1" width="500"/>
<img src="images/tension13.png" alt="plot1" width="500"/>



Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291

Clearly this element is elastic enough to break without reaching a clear plastic deformation. Values can be compared in the following table:

|                          | **NP Swab** | **OCT Industries** |
|:------------------------:|:-----------:|:------------------:|
| Tension value at failure |     34N     |         20N        |
|     Max displacement     |     17mm    |         4mm        |

After analyze the results, clearly the number 11 shows what has happened in the tension test. The first time I clamped the swabs without a rubber, damaging locally more the tip of the number 11, breaking more locally along the clamp edge and reaching a lower performance than the rest as can be seen in data.




## **Axial Torsion Fatigue Test**

Samples number 14, 15 and 16 were tested as described.  

![NPTension](images/NPTension1.jpeg)

Results are:

| **OCT sample** | **Angle Reached** |
|:--------------:|:-----------------:|
|       14       |        910º       |
|       15       |        660º       |
|       16       |        940º       |


Compared with official NP Swab:

|                   | **NP Swab** | **OCT Industries** |
|:-----------------:|:-----------:|:------------------:|
| **Angle Reached** |     710º    |        837º        |
Alfonso Parra Rubio's avatar
Alfonso Parra Rubio committed
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409

# **Second Candidate**. Resin printed Swabs.

***Raw data in csv format and python script used to read, filter and plot information can be found in the repo folders***

A couple days ago some lab friends reached us to test mechanically some of their proposed designs. On April 25th three different swabs were given. It is going to be used some tracking names to track the model and the test done as the chart shows:

<img src="images/fl/all.jpeg" alt="samples" width="600"/>

|                  | **Model/pn 1** | **Model/pn 2** | **Model/pn 3** |
|:----------------:|:-----------:|:-----------:|:-----------:|
| **Bending test** |     p1b     |     p2b     |     p3b     |
| **Tension test** |     p1t     |     p2t     |     p3t     |
|  **Axial test**  |     p1a     |     p2a     |     p3a     |

Also, each experiment will be done on two or three swabs. An additional name will be added indicating the number of the experiment.

### **Bending Test**

<img src="images/fl/bending.jpeg" alt="samples" width="600"/>


Data for the swab with PN 1, PN 2 and PN 3 is the following. It can be seen that the resin used is the most rigid of all the swabs it has been tested here. Also, the diameter of each section is bigger than swab pn2 and swab pn3.


![charts](files/fl/p1b1.png)
![charts](files/fl/p1b2.png)

PN2 has got a similar external  shape as PN1 but the resin seems more malleable in terms of bending. Even though the results doesn't show this behavior that clearly as in real life, this could be caused because when we bend with our hands the swab we clearly go over the 35 degrees that this test do. And the difference could be in the ultimate bending angle, not in the strength needed to bend 35 degrees.

![charts](files/fl/p2b1.png)
![charts](files/fl/p2b2.png)


PN3 has got clearly the thinest cross section and is the easier to bend as results shown. Almost one order of magnitude from its partners. The huge peak shown in the swab with pn3 correspond with the compression of the tip so those values should be discarted.

![charts](files/fl/p3b1.png)
![charts](files/fl/p3b2.png)
![charts](files/fl/p3b3.png)

The experiment used as usually the specific Tooling

![charts](images/fl/p2b.jpeg)

Another interesting point is that any of the swabs shown any spring back deformation after the bending test.


### **Tension Test**

The ultimate tension test shows the maturity of the process. Value are so constant for each member of the family. All the swabs shown the capacity to resist almost 3 times more load than the official NP swab. Also, the type of breakage was so fragile and immediate, with a clear defined fragile plane.

PN1 showed a maximun peak so constant in all samples of almos t 140N when deflecting 3-4mm.

![charts](files/fl/p1t1.png)
![charts](files/fl/p1t2.png)
![charts](files/fl/p1t3.png)

The breakage point in all samples were in the transition radius from the tip to the main beam.


PN2 samples , as we commented before, seems a more elastic resin that also holds a bigger plastic deformation part of the chart as it can be seen. The deformation is almost the double, than PN1 and holding 100N. Surprisingly, the breaking point happened in the transition from the big beam section to the reduced beam section in all of the samples.

![charts](images/fl/p2t.jpeg)


![charts](files/fl/p2t1.png)
![charts](files/fl/p2t2.png)
![charts](files/fl/p2t3.png)

The biggest surprise came with the PN3. The comparation with the other swabs make it seems the most brittle with a huge difference. But surprisingly, with almost the half of the mass than its partners, PN3 holds a mean of 88N, 2.3 more times than an official NP swab. Also with a fragile way of breaking, in the middle of the thin section
![charts](images/fl/p3t.jpeg)

Chamfers in all the transition locates the breakage point in a much more secure place. It is needed to avoid an uncontrolled breakage close to the tip inside a nose, so this is the safest design with difference.

![charts](files/fl/p3t1.png)
![charts](files/fl/p3t2.png)
![charts](files/fl/p3t3.png)




### **Axial Torsion Fatigue Test**

![charts](images/fl/p3a3.jpeg)



The last test is the Axial Torsion Fatigue Test. Swabs were loaded up to 10N and turned into the value decreases close to the half.

My personal expectation was that this swab, because of the type of resin were much more brittle. But results shows that absolutely no. PN3 has the record with 7 turns before breaking with is impressive. It is so remarkable that also, the breaking was so so fragile and explosive. Any other swab in this test broke so plastically but this 3 models behaves so fragile as well in this test.

PN1 behaved incredibly flexible, able to make as a mean 3 turns

| **PN** | **Angle Reached** |
|:--------------:|:-----------------:|
|       p1a1       |        950º       |
|       p1a2       |        1100º       |
|       p1a3       |        1130º       |



PN2 started to shown its more flexible behaviour. Up to 4 turns, the vale of the tension did not reduced to the half. The brittle breakage happened as mean in the 7th turn.

| **PN** | **Angle Reached until reduce tenion** | **Angle of breaking** |
|:--------------:|:-----------------:|:-----------------:|
|       p2a1       |        1420º       |  2600º       |
|       p2a2       |        1460º    |2450º       |
|       p2a3       |        1590ªº       |2580º      |

PN3 behaved simillary to PN2, having as a mean 3 turns untill reduce the tension and breaking at the 7th lap as well.

| **PN** | **Angle Reached until reduce tenion** | **Angle of breaking** |
|:--------------:|:-----------------:|:-----------------:|
|       p3a1       |        940º       |  2300º       |
|       p3a2       |        1230º    |2650º       |
|       p3a3       |        1020ªº       |2370º      |

![charts](images/fl/p3a.jpeg)