parallel cells battery monitor idea

[whatever]

just trying to think of some way to monitor a very large array of cells such as tesla pack for temperature.
Some simple way to identify if any cells temperature goes too high.
Came up with one idea that might work as follows.
Model S battery pack has a plastic cover ( clear ) over the cells as seen in photo below:


This plastic cover is directly over all the cells. All the cells emit radiation ( black body radiation) according to their temperature.
We are interested in the infrared region. This clear plastic sheet will have total internal reflection, so some of the radiation emitted will be reflected internally through ( along) the plastic. At the edge of the plastic you will be able to detect this radiation that flows through the plastic ( the internally reflected portion).
Lets say you insert an infrared sensor that picks up radiation of a temperature just below thermally damaging temperature of the cells.
I'm not sure what temp that is, lets say 75degC. or so.
Since the infrared sensor if sensitive enough should be able to detect radiation from any cell, there might be some preferred location for the sensor. Once the pack is closed there are covers in place that reduce the outside radiation entering the cells.
If the sensor detects a predetermined wavelength the pack can be isolated. It might take a few of these sensors to cover the whole pack, but the total internal reflection should mean only a few are required. Maybe one will be enough
Below is a pic to further explain the idea

 

[whatever]

just a few more thoughts:
monitoring temperature of packs, I mean monitoring individual cell temps has not really been possible, too much wiring complexity for thermocouples, the only solution thus far seems to be thermal camera on packs where all cells are visible.

I guess what i'm suggesting is to monitor using radiation of the cells and light tubes. Optic fibres are a good example of a light tube.
How to achieve diverting the radiated light ( in the form of infrared emitted from each cell) to a place where it can be easily monitored.

What if a thermal imaging camera type ccd screen ( the part of the camera that recieves the radiation) is used
( such as https://www.flir.com/products/isc0904/ or similar), 1024 by 1024 resolution, gives about 150 pixels for each cell in a 7,000 cell pack ( such as on tesla cars).
So lets say you can direct radiation from each cell to an area the size of a small flir screen. Probably not an easy feat to achieve.

IF it were possible to modify the clear plastic cover used in tesla packs to achieve this would be one solution. How to do that?
Lets say a laser heats up the clear plastic melts it slightly and on cooling the physcical properites of the plastic are changed, the area hit by the laser might be able to be manipulated such that internal reflection is used to direct radiation from a cell to a point on exterior of the plastic cover. Further the mould for the plastic could include diffraction gratings or other methods to redict radiation.

So summary: large piece of clear plastic, has light pathways manufactured into it, these pathways could be extremely small ( mm size or smaller), they all exit onto a flir type ccd screen. If these light pathways could be created by laser treatment of the plastic would be ideal. Other methods could be used but would probably be expensive and time consuming.

Did a pic to show the idea, just for 6 cells to show what the idea is , but could be used for 7,000 cells or whatever amount.


There are companies currently using optic fibres to sense temperature at multinodes or points along the optic fibre, fairly similar to this idea, technically speaking its very do-able. I've mentioned about those companies using optic fibres for temp sensing in previous posts.

 

[jonescg]

Tesla does have a patent on this. Should be easy enough to find; I know I found it once but can't remember it's title.

 

[whatever]

tesla has a patent on this method? interesting, it seems logical step to go towards.
I'll see if I can find it

 

[whatever]

did a fair bit of research on tesla patents, cant find one with same idea.
Just to think outloud a bit on the idea:
My thought is that if you were to melt some plastic material with a laser, it may be possible to make a pathway for a lightguide.
I'm guessing the melted plastic as it cools with have a certain transmittivity, rate of cooling ( as it re-solidifies) I'm guessing will
influence what the transmittivity value is. So a laser might need to be followed along by some cooling method ( gas/water etc). The same principles as used in optic fibres could then be used to direct wherever you want the light to go.
It might be necessary to have a bottom and top laser on the plastic to get some 3dimensionality to the produced light guides.
I'm thinking lasers could wizz over a large plastic sheet in very short time period and produce very complex light guides, the actual guides produced could be tiny in dimensions, maybe millimeter size or smaller.
Which leads to another thought:
would it be possible to make complex circuits for light ( infrared or other) using this principle. If the process is physically possible there would i'm sure be many different uses.,
I've done some research trying to find anyone ( researches) who have done similar to this sort of thing, as yet I"ve not found any.

Another issue on this idea is that the correct plastic would need to be used according to what wavelength light is being used,
the ir wavelength in question needs to suit a particular plastic ( eg pmma in optic fibres is transparent to ir less than 2,800nm)
I think its an interesting idea.

 

[whatever]

seems I aint so crazy afterall, some researches have achieved similar process in area of biomedical research, using exactly this idea
( well very similar) see

https://physicsworld.com/a/tiny-light-guides-could-enhance-biomedical-devices/

they managed to produce lightguies 50microns in size, now thats small, in a type of plastic ( pdms)

 

[whatever]

here is the original research paper, note the first diagram, pretty much sums it up, it is very much possible to do it.

https://www.osapublishing.org/DirectPDFAccess/32DCB5CE-9F8A-473C-8FBA80E734A13307_403007/ome-9-1-128.pdf?da=1&id=403007&seq=0&mobile=no

I suspect if the plastic tesla has used on its battery is suitable for infrared transmission that it is very much a possible solution to monitor all battery temps in real time, with addition of laser induced light guides.

here is pic from article showing a number of different aspects for the biomedical device, but microscopic lightguides are shown

 

[whatever]

a few more thoughts on the topic:
for ebikes, you could use optic fibres, there are larger diameter ones used in pool lighting etc, pretty cheap on ebay.
Say you had 14 cells, you could direct 14 optic fibres to a flir camera, and you would have real time temp monitor of all cells.

Back to tesla large number of cells pack:
there are a couple of issues to think aboutj
1. what would be min diameter light guide to use, it would depend on the sensitivity of the flir screen and the transmissivity of the light guide material, I'm guessing you could go pretty small diameter maybe under 1mm, maybe very small.
2. what to do with the output data?
an onscreen display could be used to show all cell temps at any one, simple arrangement of dots the colour of each determining the temp of each cell. It could show in real time which cells have some weakness, cells getting warmer than others would be a warning sign. on tesla could be a selection on the console screen. theres 644 cells represented below, there would be alot more dots than shown ( about 10times more).



The data from 7,000 or so cells could be very very useful. At the moment there is no way to discriminate between parallel cells, no way to know there state of health in detail. If you have a real time temperature measure of each cell, that data could be used to determine if cells are having issues.
If might even be possible to determine the internal resistance or similar type of value. If you know the temperature of a cell and you know how much current is going through it, from lab experiment on cells you would be able to get an estimate of IR .

Another issue would be lining up the plastic cover light guide with the cells, would need to reasonably accurate.
Lets say you only need 50microns or less of lightguide for this system to work, you would be able to monitor not just the cell temps but also the fuse wire temps ( if the plastic can be aligned accurately enough).

another issue in ir monitoring is emissivity, would have to be taken into account for this system, but shouldn't be a problem.



It would certainly make a could phd thesis for someone.

what would the cell screen shot show, I'm guessing you would get clusters of cells with different temps, likely it would be fairly even temp distribution, but would be some variation, my guess that variation wouldn't be scattered outliers, but clustered groups of cells with different temps. Based on my experience with lots of cells, seems manufacture process gives rise to clusters of cells with similar properties.

 

[whatever]

since tesla already has wealth of knowledge on using cameras for autopilot, it would be no problem to use camera output from a flir screen to analyse cell state of health, parallel cells would no longer be invisible. If any cell went to a critical temp point just before thermal runaway, complete pack shutdown could be implemented. It would certainly resolve fire issues to a high level.
Would the system be robust and reliable...........probably.

 

[whatever]

ok one step further:
lets say a tesla style pack is made such that the cells can be popped in and out of the pack easily, with an all cell temp monitor, you might be able to just take out bad cells and put in new ones at certain intervals or when necessary.

 

[whatever]

imagined screen output of all cells, each block of cells can be shown in more details as pic on right side.
max/min/average temps etc could be shown on same screen.

 

[whatever]

research article just showing using optic fibres for infrared temp measurements
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3028499/