Posted on 28 Comments

The Warmest Yarn is Not Qiviut – Shocking Test Results!

Hello fiber lovers!  We have all wondered what is the warmest yarn to spin and knit.  If you ask the internet what is the warmest natural fiber the consensus says Qiviut. While researching this, I couldn’t find a site that explains how this was tested so I decided to test it myself.  I went into this expecting to see Qiviut was the warmest fiber and I just wanted to see by how much was it the warmest. My testing results were shocking!

I also made a video covering this topic. If you prefer the video format you can see it here.

Results

I’m not one to build suspense so the results are directly below. These results show how well different knitted test swatches insulate compared to no insulation. The testing methods section at the end of this post goes into the full details of how the testing was done.

These are not the results I was expecting. I collected results multiple times just to make sure the results were consistent and they were. All the test swatches are as close to the same density and thickness as I could get. I did some measurements on the thickness and they are all between 4 and 4.7mm. Even if I adjust for the slight difference in thickness it doesn’t make a difference in the rankings here.

From my testing this is how I’d rank these different yarns from warmest to coolest.

  1. Alaskan Malamute
  2. Merino
  3. Alpaca
  4. Silk
  5. Angora
  6. Qiviut
  7. Rose

Paper Towel (Control)

This was included to give a control to compare against. Obviously a single sheet of paper towel isn’t a great insulator, but it is significantly better than not having any insulator. It took 3.2 times as long to heat through a paper towel than it took to heat the water with no insulator between the hot plate and the water.

Alaskan Malamute

The Alaskan Malamute is a dog that was bred to haul heavy freight as a sled dog. So it should be no surprise that the their hair when turned into yarn is extremely warm. It was my warmest sample I tested by a significant margin. The yarn only uses the dog’s soft undercoat so in addition to being very warm it is extremely soft. The sample was probably the softest of any of the samples I had and it had a beautiful halo.

The problem with this sample is this isn’t a common type of fiber which makes it hard to get and really expensive. The good news is the next warmest fiber on my list is much easier to get so let’s look at that one.

Photo by SCMW – CC BY 3.0

Merino

This is the type of wool people say to use when you want a warm sweater, and it didn’t disappoint. It suprised me that it was the second warmest yarn in my tests. I expected some of the other exotic yarns to do much better, but after rerunning the tests several times I’m convinced in my testing with my samples it is the second warmest yarn. An added bonus is this is a fairly easy to get yarn and less expensive than most of the other yarns on this list.

Photo by Fir0002 – GNU 1.2

Alpaca

Alpaca’s performed well in this test. It turns out it isn’t quite as warm as Marino, but it did get third place in my testing. Alpaca’s are native to cold environments in the Andes Mountains and their fiber is an excellent insulator. I did some looking online to see how arm Alpaca is and it was certainly classified as a warm fiber. like that my testing shows how it compares to others.

Photo by Lesbardd- CC BA-SA 4.0

Silk

I wasn’t sure how silk would perform with insulating since it is from a silk worm that uses the silk to form a cocoon to protect themselves during metamorphosis, where as wool is there to keep sheep warm. That said, silk did a good job in my tests. It took around 5 times as long for heat to permeate the silk test swatch as compared to having no isolator. Silk is also a light weight and soft fiber.

Photo by Fastily – CC BY-SA 3.0

Angora

Angora fiber is the undercoat of Angora rabbits. When I searched for the warmest natural fibers it was usually Angora and Qiviut at the top of the list. So it was a shock that this fiber came in fifth place. To be fair the fibers that placed third through fifth were all very close so you could argue that Angora was basically tied for third. Still based on what I read online I was expecting this to do better than Merino wool, but it is well behind that. On the plus side this fiber was extremely soft and had a great halo.

Photo by Ross Little – CC BY-SA 2.0

Qiviut

The internet says this fiber from a musk-ox is warmest fiber. Many places say it is eight times warmer than wool. This is where my testing completely disagrees with those statements. I found Qiviut was not even close to as warm as Merino wool or a bunch of other fibers. The sample I purchase, which is the most expensive yarn I’ve ever purchased by the ounce, was from what seems like a reputable Qiviut fiber dealer and it was very soft.

I’m still a bit in shock that that Qiviut tested so poorly in my testing. I have verified my test swatch is similar in thickness to my other samples. I would like to test another sample of Qiviut someday, but I just can’t justify that expense for this blog post right now. That said even if I find a sample that is much warmer, it would still be difficult to beat some of the other fibers tested here.

Photo by Quartl – CC BY-SA 3.0

Rose

This knitted test swatch was the worst insulator of the knitted test swatches I tested. There are certainly times where you want a beautiful knitted garment, but you don’t want it to be too warm. Maybe it’s a warm summer night, but you want to wear that shawl on the beach. In those cases Rose yarn would be a good option.

This is fiber made from rose plants and turned into yarn. It is a cellulose fiber so I didn’t expect it to be a good insulating fiber. While I didn’t test it, bamboo would probably have similar properties and is more commonly found in yarn.

Update – There is some controversy about whether rose fiber is from rose bushes. Check out this article for another point of view.

Photo by JLPC – CC BY-SA 3.0

Does Human Body Warm Impact These Results?

Testing how warm people perceive different fibers is a completely different test. I’m including this section because it’s going to be a common question and I want to try and address it.

The human body is constantly generating heat at some point the insulating properties of the clothing you are wearing keeps enough of the heat on your body comfortably warm. So while the percentages above can help you see the relative insulating properties of different fibers, you shouldn’t take these numbers to be how warm they keep you in all cases. The human body is constantly adapting and switching from warming to cooling states so there is fairly wide range of temperatures you can be comfortable. Because of that even though some fibers might be better insulators, you might not even notice a difference in many situations.

From talking with many spinners and knitters, I believe softness of a fiber likely impacts how warm people think a fiber is. The mind is powerful, and just as sugar pill placebos can help reduce certain ailments I’m pretty sure people’s perception of what a warm fiber is will help keep you a little warmer. This isn’t going to keep you warm in really harsh conditions, but if you are comparing two different sweaters on a cool fall evening perception your mindset could well make a difference. I don’t have any papers directly addressing this this type of behavior so feel free to treat this point with some skepticism if you require that kind of proof.

Some fibers keep you warmer when wet. This page does a good job of explaining why wool has this property. In certain conditions this effect will affect how warm clothing feels.

Thickness is going to matter. If you have two scarves made from the same type of yarn, but one is knitted so the scarf is thicker then it will feel warmer. Use this to your advantage by trying to knit thinner garments when you want them cooler and thicker garments when you want them warmer.

So taking all this into account if you really wanted to figure out what knitting clothing feels the warmest, a scientific way to do that would be to create simliar thinkness/gage mitten out of several different yarns. Then have a lot of people try it on and put their hand in a cold box. You shouldn’t let them see the mitten or know which one it is. Ideally you would even put some extra covering over their hand so they couldn’t feel the fiber so that perception had minimal impact on this test. Then ask them to rate the warmth of the different mittens on a scale of 1 to 10. While I’d be interested in this kind of test it was beyond the scope of this post.

Testing Methods

I want to explain how I am testing this for a few reasons. The first is I couldn’t find anyone explaining this and as I mentioned in the introduction it’s important. Another reason I’m explaining my testing method is because it isn’t a standard method. The best way to do this testing would be to use thermal conductivity meter, but these chambers seem to cost over $10K USD which is way beyond my budget. I am using a modified version of the guarded hot plate method which is one of the most commonly used methods for measuring the thermal conductivity of insulation materials. Chapter 2.1 in this document explains this method. My version is modified because I only cared about a the relative difference in the insulation properties and thus didn’t need the much more complex Poensgen apparatus described in this paper.

Here are the tools I use in my method:

  • Thermometer (I’m using a Fluke DMM with a temperature probe)
  • Scale (I use this to measure 50 grams of water)
  • Stopwatch
  • Hotplate (I’m using a solder reflow plate because it’s accurate and has a controllable temperature)

The basic procedure is I put 50 grams of room temperature water into a small beaker. Then I put the insulator (a knitted test swatch) on the hot plate set to 120 degrees Celsius. Then I put the beaker of water on the insulator and measure the time it takes to raise the water by 10 degrees Celsius while using a 3d printed cover for the beaker to hold the thermometer in a consistent position. The longer it takes to heat the water the better the knitted test swatch insulates the beaker from the hot plate.

The main issue with this test is I don’t have it calibrated to give an output of thermal conductivity in the standard units of watts per meter-kelvin. It would be nice to have that, but a system that does that is more complex and beyond the scope of this project. Instead my measurements are only useful to compare them again each other. My testing methods work because I want to find the warmest yarn and my test will tell me which of the yarns I’ve tested is the warmest relative to the other yarns.

I want to give a huge thanks to Kathryn, who goes by the the alias CraftMeHappy on Ravelry for donating many of these test swatches.

28 thoughts on “The Warmest Yarn is Not Qiviut – Shocking Test Results!

  1. FYI. Qiviut is pronounced Kee Vee Ut.

    1. Thanks!

      1. When I saw the title of this article I thought, “Oh no, he is going to say that surprisingly acrylic is the warmest fiber.” Whew…so relieved.

        1. I did want to measure acrylic just to compare. I don’t expect it to be very warm, but the results would be interesting. The problem is we couldn’t spin it ourselves and didn’t have a yarn that matched our other samples very well so it wouldn’t have been a fair comparison to include it.

  2. Thanks for doing this! I love seeing science applied to the fiber arts 🙂 One thing that I didn’t see addressed anywhere is how controlled the yarns were? Were they all spun with the same method at the same thickness? I’m not sure if this would be the right control, but I *think* maybe they should all be the same approx weight, yardage, and thickness, spun with the same method (trapping a similar amount of air). The amount of air trapped inside a yarn could definitely affect the insulation properties. Sadly, I’m not a consistent enough spinner to accomplish what I’m describing, or I’d whip up another set of rest swatches :/

    1. I did my best to make all the samples as similar as possible. The things you mentioned were made similar, but I’m sure there is some variance. I would like to do directed tests for a lot of those things by just changing one thing at a time with the same fiber to see how much that changes the insulating properties of yarn. If we find that changing how you spin or knit the yarn has huge impacts on the warmth of a fiber that would be really interesting information too.

      1. It’s “common knowledge” that the preparation of the fibers and style of spinning will affect the warmth of the end result (which is not to say it’s true, just that many have claimed it’s true). Combed vs carded fiber, woollen vs worsted spinning (although that’s a spectrum rather than bimodal). The densities of the fibers themselves matter – alpaca and angora are usually described as the warmest *per weight* because they are both hollow fibers and lighter than a similar volume of merino. Alpaca itself has 2 main types (huacaya and suri) with noticeably different qualities. And of course, the individual variations in the quality of each type of fibers would be difficult to take into account; not all merino are made equal and the term is applied to a very very wide variety of wool (what Aussies call “good merino” would barely qualify as merino at all under some standards).

  3. Go malamutes! I happen to have one and I have spun some of his fiber. I am still trying to master removing the top coat hairs though, otherwise it makes for a slightly scratchy yarn. I really am not surprised by the result, though it turns out that dog fur is not x7 as warm as wool, and angora is not 5x as warm as wool, which is what I’ve read.

    1. I want to do more testing to see if I can isolate what makes differences (maybe how the yarn is spun will matter a lot which would be interesting), but this very clearly shows that a lot of the numbers out there are either completely wrong or were came up with using some “interesting” tests. If anyone finds out where the existing numbers of 5x and 8x for Angora and Qivuit come from let me know.

  4. Wonderful experiment! As others have noted, inconsistencies in how the yarn was spun, thickness of yarn, how tightly the swatch was knit, all affect the results.

    I’m curious to know how well a given fiber keeps the water warm. I’ll try to explain. Let’s say you heat your 50g of water to temperature H, place the little beaker in a cozy, and track the rate at which the water cools off. I know this experiment would suffer from some similar downfalls as your original one with regards to the handmade nature of the swatches(cozies). One thing this second experiment removes from the equation is damage to the fibers from direct contact with the heat source. I don’t know how long it would take for these fibers to break down due to heat, or to what degree the direct contact with the heat source changes the fibers; I’m really just thinking out loud here.

    1. Thanks! I am working on a new experiment that is similar but not exactly the same as what you described.

      1. I can’t wait to see it and the results!

  5. I would love to see where Possum fit in this test. I don’t believe they have pure possum. It is usually spun with other fibers.

    1. This fiber has come up a few times. If I expand my fibers in a future test I will look into getting some opossum.

  6. Opossums and possums are different animals. Possum fibres would be Australian.

    I think something you could do as a control is have the same person spin all the fibres might help.

    Being from the arctic I’m not sure I believe the results about qiviut. I wonder if you got a blend.

    1. Thanks for the correction about possums. I wasn’t aware of this.

      The Qiviut was certainly not a blend. Just by looking at it I can see all the fiber matches. I’m looking into other testing methods that might give different results.

    2. We were in NZ once and a major invasive species there is Australian Brush-Tailed Possum. Hides are available, but the fiber itself is just too short to spin. It was being mixed with wool at a 40% possum/60% sheep ratio is my memory. Very soft, but has to be mixed. We didn’t come across ay straight unmixed fiber.

  7. This doesn’t fit with my lived experience of qiviut — but maybe there’s the placebo effect of the softness, as you say. I can see a problem with your methodology, though, if I understand it correctly. Qiviut probably gets a lot if its insulation from its loft. If you place the beaker on top you are losing more loft perhaps than in some other fibres. You may need to support the beaker so it doesn’t squash the fibre.

    1. Just looking at the fiber I didn’t see any reason that qiviut would be more susceptible to compression than other fibers. In fact some of the others seemed to compress more. That said, I agree that this is something that I should try to improve in my testing and I’ve been investigating if I can come up with a better testing apparatus.

  8. Hi Maurice. This is great fun and very interesting. Another fiber sample that is pretty common and also has been rumored to be very warm is Mohair from Angora Goats. The old wives sayings are that it is also supposed to be warmer than wool. It is another fiber that has lots of “loft”. If you ever need some raw or spun mohair I would be happy to provide you with some.

  9. Interesting! I have Siberian Huskies, I wonder how their undercoat compares to the Malamutes 🙂

    1. Our experience is that there is quite a bit of variation from dog to dog, but generally a Malamute is fuzzier. Of course, in a cold climate they all get fluffier! Don’t waste that fiber. Regular brushing is best, since that doesn’t get the guard hairs. IF its too short too easily spin on its own,mix it with something that’s longer.

  10. A comment on your testing method (also pointed out by someone else, I see):
    “Then I put the insulator (a knitted test swatch) on the hot plate set to 120 degrees Celsius. Then I put the beaker of water on the insulator . . .”

    The heat is transmitted by both conduction and convection. Placing the beaker directly on the fiber swatches will compress them, bringing the beaker closer to the heat source, thereby increasing conduction. I’d be curious about the results of running this experiment with the beaker held at a fixed height above the heat source at just the height of the idealized thickness of the swatch so there is no compression of the swatch.

    But this does point out the problems in running these sorts of experiments, namely consistency and replication. Its hard to spin and knit two items that are identical in yarn weight, thickness, and tightness of knit or weave. Maybe do 10 swatches of each experimental treatment (in this case the type of fiber) and see how those results crunch out in some summary statistics. You might compare the weight of each swatch, as well.

    My partner, Marianne, uses many different types of fibers, including dog, which was the first domesticated animal used on a regular basis for fiber. (The aboriginal peoples in the pacific northwest of north america had a distinct breed that was kept and used almost entirely for this purpose.) Softest of all the canid fibers my partner has used was from a Pomeranian. She’s also done a sweater with Samoyed, which glows iridescent white in the sun. But the nicest fiber of all (any type) was from a long haired blond dog that was an Akita-Husky (named Ganda). The fiber was beige-tan, long staple (4″-12″; undercoat up to 4″), completely waterproof (no absorption), lightweight and very fuzzy. My Ganda knit hat made from her fiber was very lightweight, completely waterproof, and incredibly warm (though the hat is now worn out after about 6 years). Our current packmate is a Malamusky named Takilma and my current hat is 40% Alpaca, 30% Malamusky, 20% Camel, and 10% Yak. Tight knit and very warm. Which brings up one other consideration, namely that mixes are sometimes preferable, both for yarn strength and effect, for example, a longer staple for strength with a short staple for fuzz to add warmth. Alpaca and Angora together would be an example. And Marianne just pointed out that a looser yarn and/or a looser knit will trap more air (up to a point) and thereby be more insulative. Think about the very thick but very loose Mohair sweaters popular in the 1950s and 60s.

    Anyway, Marianne is counting the days until her Eel arrives. Keep spinning your wheels!

    1. Thanks Steve for the great feedback. I’m actually working on a new test that tries to solve the compression issue (I still have no good feel for how much difference that causes for different fibers, but I do want to find out). My new test is basically putting a heated block of aluminum below the test swatches and seeing how long it takes to cool off. I did get some results from this, but the differences were very small. I suspect because too much of the heat from the plate is escaping by not going through the test swatch. I did insulate it, but apparently not well enough. I’m going to try some other ways of insulating around the aluminum block. If I can’t get this to work I might try your experiment suggestion, but I think not having anything to block air flow above the test swatch is better than what you suggested (at least it seems more like the case we are trying to simulate).

      I agree a large sample of test swatches would be ideal, but honestly that is perhaps too much work for me. If I get a testing apparatus that I’m happy with I might be willing to lone it to others so they can do their own testing (or accept test swatches from others). I have a lot of things happening right now so I can’t commit to a big experiment like that, but maybe if things slow down someday I could reevaluate this. There are a ton of things I’d like to test related to yarn based on feedback I got from this. I plan to at least discuss some of these if/when I do an update for this project. That might in turn spark more ideas. I have added your suggestions to the list. Thanks for the feedback!

      1. That (at least conceptually) seems to be a better approach, since it would then “measure” heat flow from both conduction and convection. Much closer to the “real world.”
        If you can put together a good standardized experimental set up, getting people to submit test swatches might be a low work way to get the replication that’s needed for statistically robust results. The hard part is limiting the number of potential variables. If there’s more than 2 or 3 analysis gets way more complicated. A quick list that might not be too onerous to measure and track:
        • size of swatch
        • weight of swatch
        • thickness of swatch
        • fiber type(s)
        Of course, replication, replication, replication!

  11. Your results don’t surprise me. I’m a very long time dog person and can tell you it doesn’t much matter what breed of dog you use as long as it’s the soft inner coat you are spinning. I would love to see the differences in something like an Afghan hound and the inner coat of double coated breeds. Sadly, I no longer have afghans so I can’t even give you yarn spun from them

  12. Which breed of angora rabbit did you use for your sample?
    Also would love to see where mohair (angora) goats rank!

    It’s really odd about the results though. Placebo effect aside, when I hold qiviut as a fiber or as a knit object my hand gets hot to the point of sweating very quickly, not at all the same as various alpaca, or merino. And outside in -40 the qiviut hat or headwrap is the go to, vs any other hat, as there is a noticeable difference in trapped warmth around ears. Rabbit is a close second.
    Could it be an extreme cold thing, or an exposure thing, I wonder?
    Watch out, I’m going to be standing out in the cold in a month or so with a sensor on my head and sending results….

    Another question:
    Many times straight up qiviut or angora rabbit are thinner closer to lace weight. I wonder if thicker yarns of those fibers aren’t as effective?

    1. Thanks for the feedback. I’m hoping to be able to do more testing in the future. Perhaps with more samples.

      Sorry, I’m not sure what type of Angora rabbit fiber was used. My wife purchased the fiber many years ago at a fiber festival and it didn’t have any additional info about this with the fiber.

Leave a Reply

Your email address will not be published. Required fields are marked *