EEVblog Electronics Community Forum
Products => Thermal Imaging => Topic started by: CelXar on January 15, 2024, 03:10:05 am
-
I'm not that familiar with thermal cameras but I've been doing a lot of research, looking at YouTubes, articles, etc and I think I've decided between a Hikmicro B10, Guide PC210, or the iphone Infirary P2 Pro model to at least pilot a test. But I have a very specific use-case I'm looking for that I haven't seen covered by testing (most are focused inside buildings or PCB signatures), so I wanted to check if anyone had any feedback or experience.
I work in Dam Safety and operate a number of large concrete spillway structures. We know have various delaminations through concrete sounding tests (drummy areas); however, I wanted another tool in our arsenal to better define areas of concrete delamination and potentially identify new areas. I'm guessing the ideal time to scan might be either transitioning to dusk or early dawn. Either that or perform a water surface test by placing water on the surface and looking for differences in evaporative capacity.
I'm not interested in accurate temperatures, just thermal differential signatures where delaminations should show a different heat signature than intact concrete. Does anyone have any experience with a particular (affordable) model that would work well outside for scanning concrete delaminations that could pickup relatively small temperature differentials, and from a distance? Is a 256x192 (or similar) IR resolution adequate for this purpose?
-
Generally speaking, safety usually has a decent budget and can afford to dedicate a person or two for specific training. That may not be your case. If so, I'd suggest a thermography class or two. Understanding the fundamentals of emmisivity is key to this type of work. Can also provide you a better understanding of the minimum piece of equipment you need. Careful as some of the training classes are also marketing who try to sale you on their brand of equipment.
Camera selection basically boils down to 1) how big of an area to you want to see at one time; 2) what is the smallest feature in that single image; 3) how far away can you be (test setup); 4) sensitivity of camera; 5) how much software automation do you want. [ 1) Field of View; 2) Instantaneous Field of View; 3) Range / Configuration; 4) 3D Noise; 5) Analysis ]
If you are really close then a single lens and small array may be sufficient. If you will be looking at a lot of different dams and the distance will vary then going with a system with interchangeable lenses may be required.
Youtube has 'some' good content, but filtering out the snake oil salesmen and 'ghost hunter' can be a challenge for the uninformed. Suggest you data mine papers on the topic and then locate the references and read those. Keep going to figure out what equipment they have used and the techniques utilized. Figure out which authors are referenced the most. You want to look for places like Optical Engineering(OE) and SPIE. More toward OE as it's peer reviewed. Research Gate is a good and usually free starting point. It's unfortunate but OE and SPIE are behind a pay wall for most of the papers.
https://www.researchgate.net/publication/317620823_THERMAL_IMAGING_OF_CONCRETE_DAM_SURFACES_TO_SUPPORT_THE_CONTROL_OF_THE_EVOLUTION_OF_PATHOLOGIES (https://www.researchgate.net/publication/317620823_THERMAL_IMAGING_OF_CONCRETE_DAM_SURFACES_TO_SUPPORT_THE_CONTROL_OF_THE_EVOLUTION_OF_PATHOLOGIES)
https://www.researchgate.net/publication/328958896_Infrared_thermography_for_detecting_defects_in_concrete_structures (https://www.researchgate.net/publication/328958896_Infrared_thermography_for_detecting_defects_in_concrete_structures)
quick search results above ... also there appears to be a "International Conference on Fracture Mechanics of Concrete and Concrete Structures" ... no idea of quality of the conference but where there is one then there are likely others.
Interesting problem. But hey, I'm a geek!
-
Oh I forgot ... look for NDT (Non Destructive Testing) books. They usually have a section or two on structure analysis. I've not looked but based on the quick papers search I bet there is something out there.
-
Thanks for the input. In terms of budget, yeah we do have a very large budget, in theory, but the way "Equipment" is budgeted for is woefully inadequate. It's a pain to justify even buying a $150 computer monitor due to the way equipment is very tightly locked down. I was going to try something paid out of pocket first (hence looking at <$400 options), and if it works or shows potential then put in a larger budget request for perhaps some expensive FLIR or something. But I wanted to come in armed with some demonstration that it could be a feasible condition assessment tool.
Some cursory reading beforehand seemed like a mixed bag, or not very useful (particularly because even just discoloration can cause issues and the concrete surfaces have varying levels of discoloration), but I haven't ruled it out. It seems like in theory, it could work, but in practice it might not work as well as I'd like no matter what quality of camera is used.
I'll take a look at the papers and see what they did. 1) In terms of area to see, I was thinking about <10 foot square, give or take. 2) Smallest feature being probably be less than an eight of an inch, but not sure. 3) Plan would generally "walk" the structure at roughly normal eyesight looking down or 45-degree, so about 6-10 feet viewing distance, 4) Sensitivity not sure, I think it would need to pickup pretty small differentials given very minor millimeter-sized delaminations might be hard to detect without strong enough differential signatures, 5) I suppose ideally automation would be to setup a standard "average" temperature baseline from a particular snapshot, and then highlight small differentials (say about 1F). For example, this was some work done with hammer soundings, but being able to replicate this with some thermal differentials in heat dissipation where delaminations might exist (i.e. the conductivity of delamination concrete should be different than intact concrete, even if it's a small difference in conductivity) would be what I'd be looking for (see attachment)
As for other NDT, we did do GPR and impulse-echo, but it was pretty poor results that didn't match concrete core tests. They also aren't really adequate for delaminations. We're looking into doing impulse-response and some higher frequency GPR as well, that's something in the works but that would be contracted out. I'll see if any books have information on thermal-application as a NDT method.
But yeah thanks again, I'm definitely out of my league here as a civil engineer trying to dabble in technology that I only have a cursory understanding of. I've tried talking to other owners but no one really had any idea or tried it themselves, and it is a very specific niche (mostly untested) application to be sure.
-
It may be a 'try one out and see what works' sort of situation. A 10 square foot area is not all that big, so I wouldn't expect resolution needs to be extremely high, but if you want differentiation of a millimeter scale feature at just under a 1 square meter field of view... that's actually higher resolution than most cameras available, period. Now if these features have a significant effect on temperature (maybe a tenth of a degree or more), then they shouldn't be too hard to spot, but you'd have to capture them with motion, closer inspection, or super resolution techniques, which is going to need different (longer) procedure and maybe different hardware.
Another thing that comes to mind with a walk-around kind of inspection is framerate - lots of cameras are locked to 9 Hz framerate because of export restrictions and they would probably be completely unusable in that kind of workflow because of how slow you'd have to move waiting for it to update. Then if you're doing QA/safety sort of work, I would imagine built in annotations or report making may be critical, which is a feature usually of the higher end stuff.
But as you seem to know, I'm not sure how much this has been done before and the answer to find a good set of parameters to look for in a device to give you some confidence it works is probably just to try it out. If looking for an 'experiment' camera on a budget, I'd prioritize sensitivity over all else. You can take the framerate or field uniformity hit if you can get enough fine temperature detail to see the features you want (this will also let you know if it's even possible to see), and the size of the inspection area and distance to target should be largely dependent on the lens configuration, so for a proof of concept may not be important. While aiming for a higher resolution for your test camera is probably desirable, it may not be that important since you can move the sample to where you need it for the resolution you have, and use that to predict what minimum resolution for the 'real deal' tool you'll need for the inspection at the distances you're looking for.
Another potential complication - with better lenses and cameras, the depth of being in focus may not be that deep, so while it should be fine for flat surfaces where you can stay relatively constant distance from your target, when inspecting at varying distances or on curved features, there may be some additional refocusing that's required as you go - it's unusual to find an autofocus thermal lens.
-
You mention "contracted out" for the GPR. I would assume that you may be looking for that as an option on the thermography side? Finding a qualified person may be a challenge but I'd bet somebody else in the dam industry has utilized a thermographic technique. A little deeper digging and saw flash (pulsed) thermography has been used on lots of structures to find voids/cracks which I'd assume lead to what you are calling de-lamination. Might not be the best as what I've seen you need to be up close to the object. The longer distance and larger surface stuff is big money.
DaJMasta makes good points. A cheap camera will be hamstrung with a slow frame rate, usually a very poor lens, and I'd add questionable sensitivity. Cheap cameras are good for looking at general area and large delta temperatures. You might be able to find a deal or two on eBay but normally a 'new' camera on the bottom end of what you are looking for is likely starting in the 500-800 and up. Just making an image is fairly cheap. It's the scientific back end that creeps the price up.
It's an interesting issue and thought I about driving out to a dam (live near several TVA reservoirs) ... but not risking jewelry that connects at the wrists and a private meeting with the FBI if I went around taking infrared pictures of a dam!!
-
The most important consideration is when you make your observations. As the sun comes up, a delaminated area should warm up faster.
Long before thermal cameras were consumer items early morning was when airborne geological thermal imaging was done as it provides the greatest contrast in soil conductivity. In your case the conductivity is the same, but the thermal resistance of the crack will result in those areas warming up faster. Time lapse imaging should give the best results as you can average the temperature changes from frame to frame.
I have a dissertation on GPR inspection of concrete structures done in Sweden. If that is of interest, I'll dig it up and provide the citation.
Have Fun!
Reg