I've been doing online research as well as some tinkering and was planning on building a prototype to demonstrate the first ever vacuum balloon, but I'm running into issues with expenses and time. I believe I've identified 2 approaches using well known materials that should work but one in particular that could be pulled off by a garage tinkerer with extra time and money to spare on the project.
Along the way I also started experimenting with creating foams using a technique I've basically invented as far as I can tell. I can't find any literature on it. I've gotten mixed results with it and am just not sure if it will ever work at least without being done properly in a lab setting. The approach has a lot of promise and I'll explain why.
There's a lot to go into on this subject. I've written about vacuum balloons before so if this is a new concept for you, you should give it a read.
I'm human so some of this work could have errors in it, but I have done experiments to test my theory and gotten interesting results. I have measured weight reduction in some of my designs and I have accurately predicted the results in cases where I could measure properly. That gave me a lot of hope to continue on at first but it's just a lot of work and I went way over budget early on. I can't keep pouring money into the project anymore and it hurts to say that because some of the results are so interesting. Also, life gets's busy and I can only tinker for so long.

Shapes

The best shape is a sphere because you need to withstand the atmospheric pressure outside the balloon pushing in at about 14 psi. For the same reasons we build bridges with arches, the sphere is the best shape for this because it will spread the forces out evenly. It becomes a matter of having a material that can withstand the compressive forces and in the case of non-uniformity (which to some degree is always going to be present) shear forces. Of course, the material also needs to be lightweight or it will never lift. Many sources will erroneously tell you no such material exists, but this isn't true. In theory, there are multiple materials that would probably work but the issue starts to become the total size of the balloon (and defects.) You could make it out of glass, but the balloon would have to be incredibly large and would be insanely prone to shattering and that's even if it was made defect free so there's really no point in trying normal glass. This is where choosing your materials is key so that you don't waste your time.
The volume of a sphere is V = 4/3πr^3
To calculate the buoyant force of lift at atmosphere you can simply multiply the volume by 1.29 kg/m3 and that will give you the amount it can lift in kg. Simply multiply by 2.2 for conversion to get the number in pounds. This formula was derived from the formula below.
​
https://preview.redd.it/56czvmdcuh4b1.png?width=516&format=png&auto=webp&s=31538f933c110d46a7d9f66af5fc8fca864bbd14
The 1.29 kg/m3 is the fluid density of atmosphere and I simply removed the acceleration of gravity to show the force in units of pure weight rather than in Newtons. It's a simple calculation and understanding it is key to helping you design the vacuum balloon.
Now that you understand how to calculate the lifting force of vacuum in a sphere you can run a bunch of numbers and see for yourself that the lifting force is very small below radius 1 and grows exponentially above radius 1. This means it will be exceptionally hard to build a working vacuum balloon below radius 1 but unfortunately there are limitations to building large structures as well. Usually you want a prototype to be simple and cheap, not experimental in and of itself. This means the first demonstrated vacuum balloon will likely be about 2 meters in diameter or about 6 feet. It also means a vacuum balloon of very large proportions would potentially have incredible lifting force.
Now that you understand the relations between size and lifting force all you need to do is calculate the volume of the envelope of the spherical balloon. This is done by simply calculating the volume of a sphere of the size of the envelope and then subtracting that by the volume of the inner void. The difference is the volume of your envelope and you can easily calculate the weight of your envelope by multiplying the density by the volume. If you do this while calculating the lifting force and plug different numbers in you can easily see how the ratio of weight to volume works. You can also see how the density influences this and even can compare the volume of different shapes if you really want to just to see how much better a sphere really is than perhaps a square.
It's very important to point out that one of my biggest lessons in building prototypes is that there can't be any defects. I originally was making hemispheres and trying to join them together before pumping down to vacuum and every time there was a failure it was at the meeting of the two hemispheres. One solid piece seems to be necessary. It's conceivable that two hemispheres can be joined and bonded to become one solid piece free of defects, but I unfortunately did not have the materials to do this. I did do some experiments and found that you can reinforce this area with lightweight bamboo if necessary. However, these were small preliminary designs and I'm not confident that would scale well.
It's worth noting that the next best shape is a cylinder with hemispheres on each end. Basically a tic tac shape. It's only worth attempting this shape if you have reasons to from a manufacturing perspective. For example, I played around with the idea of making a foam sheet and then rolling it into a cylinder before it set rather than attempting to cast a foam hemisphere. It only makes sense if you are attempting a volume too large to pull off as a sphere for practical reasons (like it would't fit in garage or won't caste evenly.) Because it still needs hemispheres it's a design best left for after demonstrating a spherical design.

Materials

I dive into the use of aerogels and xerogels in the article referenced above. The purpose of these foam materials is because when engineered properly they retain a lot of their strength but lose a lot of their weight which actually increases their strength to weight ratio and that's exactly what we need to make this work. There is no material in bulk form worth pursuing for this design. You absolutely have to use a foam material. Even if you could pull it off using glass or beryllium, it's just not practical even for demonstration purposes. During my search I found the most attractive material in the bulk to be polycarbonate. It's still not worth trying in bulk form, so I invented a way to make a foam out of it. Polycarbonate is lighter and stronger than glass. Nobody has ever made an aerogel out of it that I'm aware of. I did not image my foam because I'm not doing this work in a sophisticated lab, but I can say fairly confidently that it's about 75% porosity. That's impressive, but I suspect that a lot of the bonding is weak and there's defects, but in my defense I used an insanely primitive and low tech technique.
There are two well known foams we all have access to that in theory should work. Styrofoam and polyurethane.
I understand that may cause you to sigh in disbelief. After all, polyurethane was invented in the 1930's at IG Farben and styrofoam in the 1940's so they are not only old but very ubiquitous. I should also point out that aerogel was invented in the 1930's and was once mass produced by Monsanto. None of these materials are new.
I used the given compressive and shear strengths published by a local styrofoam manufacturer to identify some common commercial grade foams that are very light weight that should work in theory if there's no defects. I tried working with them to have some custom shapes made, but they unfortunately are limited to 4 feet for one of the dimensions of their die blocks. This is very problematic even if we knew how to fuse two styrofoam hemispheres together. I'm not going to say it's impossible, but it makes pulling it off more challenging. I did do some experiments with small 1 foot diameter styrofoam hemispheres that are commonly available and managed to measure a weight reduction before it imploded. Anybody can replicate these experiments. I expected it to fail because the thickness was less than 1 inch. I found the best design was to nest two of these styrofoam spheres within each other but with the orientations opposing so that the point of failure for the outer sphere was across the strongest points of the inner sphere. This should create a perpendicular crossing of the hemispheres of the inner and outer shells. This is also where I tried some glues. Gorilla glue works best and sure enough it's a polyurethane. I was so impressed by it that I switched over to attempting polyurethane designs for the sphere.
I found a polyurethane foam used in boating that is only 2lb/ft3 which is very impressive. It also boasts a compressive strength of 38 psi. I figure that means half an inch of this stuff would be able to handle 19 psi theoretically. That's 5 psi above the 14 psi we need for our vacuum balloon. It's not a lot of room for error, but it works in theory.
What I like about polyurethane is that you can fairly easily make custom shapes with it and DIY. I experimented with a few different techniques and can say that you need this foam to be open to the air to set properly, but it does take on conformal shapes fairly well. The best method I found to make a hemisphere out of it was to actually blow up a rubber balloon and fit that snug into a styrofoam sheet for support and then pour the polyurethane foam onto it and let it set. You can then use cutting tools to clean up the extra material. This method works, but the cutting is a pain as I did it by hand. Precision will likely be necessary to properly join the two hemispheres and I learned this the hard way when I tried to join them. A more precise way to form the hemispheres I found was to buy plastic hemispheres and coat them in wax (to make removal of the polyurethane easier.) This is far more expensive than the balloon but gives more precise results. You can find people selling these in sizes up to 6 feet but it will get pricey. It's worth mentioning that I had a hard time removing the set polyurethane from the plastic even with a wax coating (which I also verified experimentally is the least sticky thing to use) so I'm not sure it's even the best approach. I've tried reaching out to polyurethane component manufacturers but so far no response. I'm sure outsourcing this would remove a lot of headaches, but also be very expensive for such a custom piece.
Just to highlight why I think this commonly available polyurethane foam is promising I want to calculate a 1 meter radius sphere of one half inch thickness to show that it should work in theory. Of course, this means no defects including the joining of the two hemispheres which is still a problem to solve but it's possible gorilla glue and precision would solve it. Maybe a DIY'er with their own CNC may want to give it a shot.
Using the volume of sphere formula given above we see that the volume of 1 meter radius is 4.187m3. The volume of a sphere of 1 meter minus 1/2 inch is 4.0295 m3. The buoyant lift of that is 11.44 lbs. The difference in volume (to find the volume of the polyurethane used) is .1575 m3 or 5.56 ft3. At a density of 2 lbs/ft3 that gives a weight of 11 lbs of polyurethane. That's less than the 11.44 lbs of lift.
I know what you're probably thinking. How does it hold vacuum? It's true that polyurethane and styrofoam are not expected to hold vacuum (I actually did find experimentally that styrofoam does hold partial vacuum for a few hours after it's shrunk much like the LANL aerogel) but you can simply wrap the sphere in plastic to hold vacuum. I planned on experimenting with dip coatings, but for experimental purposes I came up with a very clever design that I will explain later. Just know that the plastic doesn't have to be very thick to hold vacuum so it's very much within the range of possibility to coat the sphere in a thin plastic layer at less than .44 lbs. Plastic is very dense, but we are talking about literally a few mils of material. This is also why I roll my eyes at people who mock me for attempting a design with materials that don't hold vacuum. You are not limited to materials that hold vacuum for your design when you can simply add a layer for that later.

Experimental Set Up

I initially bought one of those vacuum chambers made out of a large steel pan and thick acrylic. Mechanical pumps are easy to find and relatively cheap. Mine came with the chamber. However, I quickly found it wasn't big enough and attempting to build a larger one looked costly. This is where I got clever and shocked myself with a very cheap set up that actually works. I simply bought regular large sized vacuum bags designed for storing cloths because they have a clever little self sealing mechanism that traps the vacuum. These bags are not meant for actual vacuum with a mechanical pump so I wasn't sure how it would work. I also had to find a way to rig it all up. As funny as it sounds my solution was to take the nozzle of an empty plastic bottle that happened to fit onto the hose and then I cut a piece of EDPM rubber to cover the end meant for the bottle and put a small slit in the center for air to move through. I then pushed this into the self sealing part of the vacuum bag and it actually creates a seal and pumps down! And when you remove the pump it self seals!
I found I sometimes had issues with pumping down properly and solved this by using a metal straw that I placed inside the bag near the seal and directed towards the sphere to act as a channel. Once again, to my surprise this works very well.
So, I then disassembled my original steel pot vacuum chamber and used the parts along with some parts I had to buy online to rig the pressure gauge into the system so that I could verify how much vacuum I was achieving. I'm a bit proud of this DIY set up because it works so well.
In order to properly record your results you must weight the vacuum bag and the metal straw as well as your experimental sphere before vacuuming. Then vacuum it down and pay attention to the gauge. If your design is not very good it may implode before achieving full vacuum. That's okay. You can actually measure a weight reduction without reaching the full vacuum. "Full" vacuum in this case is actually what is known as low vacuum. Low vacuum is all you need for a vacuum balloon to work as you have effectively removed most of the air and it's not necessary to reach medium or high vacuum.
This set up was for spheres of only 1 foot diameter and I don't think there are bags large enough for 6 foot spheres. However, my plan was to use a heat gun to stitch a bunch of the bags together to make it work. It's dirty but once again it should work theoretically. I was also planning on using a heat gun to section off portions of the bag to seal it around the sphere and cut off excess material but that part is really only necessary if you are about to achieve lift. I imagine it's possible once you've proven you can make a structure strong enough and light enough for lift that a better technique would be to incorporate a valve and find a way to dip coat the sphere to seal it. I never got this far.

A Potential New Approach To Foam

I mentioned experimenting with making foams and identifying polycarbonate as good material to turn into a nano foam. I use the term nano foam because aerogel wouldn't be technically correct. They are both nano foams. The aerogel is made using gel. This approach doesn't. It's very low tech and dirty. I theorized I could use the fact that polycarbonate is a thermoplastic to my advantage and mix it as a powder with another material that can withstand it's glass transition temperature but is also easily soluble in water. So, I found some polycarbonate powder (first American apparently to buy it) and mixed it with some ordinary table salt then put it in the oven. I know this sounds ridiculous. Then I washed the sample after it cooled in the sink and dried it with paper towels. Then I soaked it in rubbing alcohol and dried that with paper towels. Then I let it sit overnight to fully evaporate if it's a big sample. Then I weighed it. When I mix the powder in a 1:1 ratio by weight the sample after washing it weights exactly half of when I started without losing any volume. So I washed out all of the salt. But, that's not all. Because this method is basically sintering the particles together, it already had lots of air pockets in it to begin with. I attempted to make a one cubic inch sample to measure the density and it's not the most precise but the density is roughly 4.7 g/in3 which is about a quarter of the density of bulk polycarbonate. This means it's porosity is about 75%. It's not he 90-99.99% of commercial aerogel, but I personally find the initial results surprising. There's a lot of ideas I have to tweak this including playing with the mix ratio, grain size, uniformity of the particles, and aerating the powder. What I find very interesting about this technique in general is that it actually would work with anything that can be sintered including other thermoplastics, ceramics, glasses and metals. This means this approach could be used to make porous metals or even metal nano foams.

The 2009 analysis of the metal sphere UFO

I've recently been made aware of the 1994 spherical UFO that Steve Colbern published a report on in 2009. A few things stand out to me as someone who has been actively working on vacuum balloons and ways to make porous metals. First, it looks like two hemispheres nested inside each other exactly as I describe was my best approach to making a vacuum balloon based off of experimental results. Second, the sphere is presumably hollow. Third, the report clearly states that the sample analyzed was a porous metal with nanostructures present. A hollow porous shell with nested hemispheres of opposing orientation is exactly what I would expect a vacuum balloon to look like. There are ways to use my technique on titanium to make it porous although I haven't done so experimentally because it's melting point is very high. Materials other than salt could be used but even if salt was used it would be interesting because it would vaporize at the glass transition temp of titanium which actually might help make it more porous. I do believe Na and Cl impurities were present in the sample according to the report. Perhaps one could experimentally recreate this sample using this method (minus the isotopes.)

Crowdsourcing

If anybody wants to crowdsource the work on this with me I'm open to it. Also, if people are open to crowdfunding the research I'm open to that as well. Either way, it's up on the internet now. Maybe 10 years from now somebody as crazy as me will pick up where I left off. I might return to this at a later date, but without help I think I need to take a break.

I've been doing online research as well as some tinkering and was planning on building a prototype to demonstrate the first ever vacuum balloon, but I'm running into issues with expenses and time. I believe I've identified 2 approaches using well known materials that should work but one in particular that could be pulled off by a garage tinkerer with extra time and money to spare on the project.
Along the way I also started experimenting with creating foams using a technique I've basically invented as far as I can tell. I can't find any literature on it. I've gotten mixed results with it and am just not sure if it will ever work at least without being done properly in a lab setting. The approach has a lot of promise and I'll explain why.
There's a lot to go into on this subject. I've written about vacuum balloons before so if this is a new concept for you, you should give it a read.
I'm human so some of this work could have errors in it, but I have done experiments to test my theory and gotten interesting results. I have measured weight reduction in some of my designs and I have accurately predicted the results in cases where I could measure properly. That gave me a lot of hope to continue on at first but it's just a lot of work and I went way over budget early on. I can't keep pouring money into the project anymore and it hurts to say that because some of the results are so interesting. Also, life gets's busy and I can only tinker for so long.

Shapes

The best shape is a sphere because you need to withstand the atmospheric pressure outside the balloon pushing in at about 14 psi. For the same reasons we build bridges with arches, the sphere is the best shape for this because it will spread the forces out evenly. It becomes a matter of having a material that can withstand the compressive forces and in the case of non-uniformity (which to some degree is always going to be present) shear forces. Of course, the material also needs to be lightweight or it will never lift. Many sources will erroneously tell you no such material exists, but this isn't true. In theory, there are multiple materials that would probably work but the issue starts to become the total size of the balloon (and defects.) You could make it out of glass, but the balloon would have to be incredibly large and would be insanely prone to shattering and that's even if it was made defect free so there's really no point in trying normal glass. This is where choosing your materials is key so that you don't waste your time.
The volume of a sphere is V = 4/3πr^3
To calculate the buoyant force of lift at atmosphere you can simply multiply the volume by 1.29 kg/m3 and that will give you the amount it can lift in kg. Simply multiply by 2.2 for conversion to get the number in pounds. This formula was derived from the formula below.
https://preview.redd.it/su8ya13m0h4b1.png?width=516&format=png&auto=webp&s=d7db2ab0b6678d6abc010f1a0a2cf6020633b344
The 1.29 kg/m3 is the fluid density of atmosphere and I simply removed the acceleration of gravity to show the force in units of pure weight rather than in Newtons. It's a simple calculation and understanding it is key to helping you design the vacuum balloon.
Now that you understand how to calculate the lifting force of vacuum in a sphere you can run a bunch of numbers and see for yourself that the lifting force is very small below radius 1 and grows exponentially above radius 1. This means it will be exceptionally hard to build a working vacuum balloon below radius 1 but unfortunately there are limitations to building large structures as well. Usually you want a prototype to be simple and cheap, not experimental in and of itself. This means the first demonstrated vacuum balloon will likely be about 2 meters in diameter or about 6 feet. It also means a vacuum balloon of very large proportions would potentially have incredible lifting force.
Now that you understand the relations between size and lifting force all you need to do is calculate the volume of the envelope of the spherical balloon. This is done by simply calculating the volume of a sphere of the size of the envelope and then subtracting that by the volume of the inner void. The difference is the volume of your envelope and you can easily calculate the weight of your envelope by multiplying the density by the volume. If you do this while calculating the lifting force and plug different numbers in you can easily see how the ratio of weight to volume works. You can also see how the density influences this and even can compare the volume of different shapes if you really want to just to see how much better a sphere really is than perhaps a square.
It's very important to point out that one of my biggest lessons in building prototypes is that there can't be any defects. I originally was making hemispheres and trying to join them together before pumping down to vacuum and every time there was a failure it was at the meeting of the two hemispheres. One solid piece seems to be necessary. It's conceivable that two hemispheres can be joined and bonded to become one solid piece free of defects, but I unfortunately did not have the materials to do this. I did do some experiments and found that you can reinforce this area with lightweight bamboo if necessary. However, these were small preliminary designs and I'm not confident that would scale well.
It's worth noting that the next best shape is a cylinder with hemispheres on each end. Basically a tic tac shape. It's only worth attempting this shape if you have reasons to from a manufacturing perspective. For example, I played around with the idea of making a foam sheet and then rolling it into a cylinder before it set rather than attempting to cast a foam hemisphere. It only makes sense if you are attempting a volume too large to pull off as a sphere for practical reasons (like it would't fit in garage or won't caste evenly.) Because it still needs hemispheres it's a design best left for after demonstrating a spherical design.

Materials

I dive into the use of aerogels and xerogels in the article referenced above. The purpose of these foam materials is because when engineered properly they retain a lot of their strength but lose a lot of their weight which actually increases their strength to weight ratio and that's exactly what we need to make this work. There is no material in bulk form worth pursuing for this design. You absolutely have to use a foam material. Even if you could pull it off using glass or beryllium, it's just not practical even for demonstration purposes. During my search I found the most attractive material in the bulk to be polycarbonate. It's still not worth trying in bulk form, so I invented a way to make a foam out of it. Polycarbonate is lighter and stronger than glass. Nobody has ever made an aerogel out of it that I'm aware of. I did not image my foam because I'm not doing this work in a sophisticated lab, but I can say fairly confidently that it's about 75% porosity. That's impressive, but I suspect that a lot of the bonding is weak and there's defects, but in my defense I used an insanely primitive and low tech technique.
There are two well known foams we all have access to that in theory should work. Styrofoam and polyurethane.
I understand that may cause you to sigh in disbelief. After all, polyurethane was invented in the 1930's at IG Farben and styrofoam in the 1940's so they are not only old but very ubiquitous. I should also point out that aerogel was invented in the 1930's and was once mass produced by Monsanto. None of these materials are new.
I used the given compressive and shear strengths published by a local styrofoam manufacturer to identify some common commercial grade foams that are very light weight that should work in theory if there's no defects. I tried working with them to have some custom shapes made, but they unfortunately are limited to 4 feet for one of the dimensions of their die blocks. This is very problematic even if we knew how to fuse two styrofoam hemispheres together. I'm not going to say it's impossible, but it makes pulling it off more challenging. I did do some experiments with small 1 foot diameter styrofoam hemispheres that are commonly available and managed to measure a weight reduction before it imploded. Anybody can replicate these experiments. I expected it to fail because the thickness was less than 1 inch. I found the best design was to nest two of these styrofoam spheres within each other but with the orientations opposing so that the point of failure for the outer sphere was across the strongest points of the inner sphere. This should create a perpendicular crossing of the hemispheres of the inner and outer shells. This is also where I tried some glues. Gorilla glue works best and sure enough it's a polyurethane. I was so impressed by it that I switched over to attempting polyurethane designs for the sphere.
I found a polyurethane foam used in boating that is only 2lb/ft3 which is very impressive. It also boasts a compressive strength of 38 psi. I figure that means half an inch of this stuff would be able to handle 19 psi theoretically. That's 5 psi above the 14 psi we need for our vacuum balloon. It's not a lot of room for error, but it works in theory.
What I like about polyurethane is that you can fairly easily make custom shapes with it and DIY. I experimented with a few different techniques and can say that you need this foam to be open to the air to set properly, but it does take on conformal shapes fairly well. The best method I found to make a hemisphere out of it was to actually blow up a rubber balloon and fit that snug into a styrofoam sheet for support and then pour the polyurethane foam onto it and let it set. You can then use cutting tools to clean up the extra material. This method works, but the cutting is a pain as I did it by hand. Precision will likely be necessary to properly join the two hemispheres and I learned this the hard way when I tried to join them. A more precise way to form the hemispheres I found was to buy plastic hemispheres and coat them in wax (to make removal of the polyurethane easier.) This is far more expensive than the balloon but gives more precise results. You can find people selling these in sizes up to 6 feet but it will get pricey. It's worth mentioning that I had a hard time removing the set polyurethane from the plastic even with a wax coating (which I also verified experimentally is the least sticky thing to use) so I'm not sure it's even the best approach. I've tried reaching out to polyurethane component manufacturers but so far no response. I'm sure outsourcing this would remove a lot of headaches, but also be very expensive for such a custom piece.
Just to highlight why I think this commonly available polyurethane foam is promising I want to calculate a 1 meter radius sphere of one half inch thickness to show that it should work in theory. Of course, this means no defects including the joining of the two hemispheres which is still a problem to solve but it's possible gorilla glue and precision would solve it. Maybe a DIY'er with their own CNC may want to give it a shot.
Using the volume of sphere formula given above we see that the volume of 1 meter radius is 4.187m3. The volume of a sphere of 1 meter minus 1/2 inch is 4.0295 m3. The buoyant lift of that is 11.44 lbs. The difference in volume (to find the volume of the polyurethane used) is .1575 m3 or 5.56 ft3. At a density of 2 lbs/ft3 that gives a weight of 11 lbs of polyurethane. That's less than the 11.44 lbs of lift.
I know what you're probably thinking. How does it hold vacuum? It's true that polyurethane and styrofoam are not expected to hold vacuum (I actually did find experimentally that styrofoam does hold partial vacuum for a few hours after it's shrunk much like the LANL aerogel) but you can simply wrap the sphere in plastic to hold vacuum. I planned on experimenting with dip coatings, but for experimental purposes I came up with a very clever design that I will explain later. Just know that the plastic doesn't have to be very thick to hold vacuum so it's very much within the range of possibility to coat the sphere in a thin plastic layer at less than .44 lbs. Plastic is very dense, but we are talking about literally a few mils of material. This is also why I roll my eyes at people who mock me for attempting a design with materials that don't hold vacuum. You are not limited to materials that hold vacuum for your design when you can simply add a layer for that later.

Experimental Set Up

I initially bought one of those vacuum chambers made out of a large steel pan and thick acrylic. Mechanical pumps are easy to find and relatively cheap. Mine came with the chamber. However, I quickly found it wasn't big enough and attempting to build a larger one looked costly. This is where I got clever and shocked myself with a very cheap set up that actually works. I simply bought regular large sized vacuum bags designed for storing cloths because they have a clever little self sealing mechanism that traps the vacuum. These bags are not meant for actual vacuum with a mechanical pump so I wasn't sure how it would work. I also had to find a way to rig it all up. As funny as it sounds my solution was to take the nozzle of an empty plastic bottle that happened to fit onto the hose and then I cut a piece of EDPM rubber to cover the end meant for the bottle and put a small slit in the center for air to move through. I then pushed this into the self sealing part of the vacuum bag and it actually creates a seal and pumps down! And when you remove the pump it self seals!
I found I sometimes had issues with pumping down properly and solved this by using a metal straw that I placed inside the bag near the seal and directed towards the sphere to act as a channel. Once again, to my surprise this works very well.
So, I then disassembled my original steel pot vacuum chamber and used the parts along with some parts I had to buy online to rig the pressure gauge into the system so that I could verify how much vacuum I was achieving. I'm a bit proud of this DIY set up because it works so well.
In order to properly record your results you must weight the vacuum bag and the metal straw as well as your experimental sphere before vacuuming. Then vacuum it down and pay attention to the gauge. If your design is not very good it may implode before achieving full vacuum. That's okay. You can actually measure a weight reduction without reaching the full vacuum. "Full" vacuum in this case is actually what is known as low vacuum. Low vacuum is all you need for a vacuum balloon to work as you have effectively removed most of the air and it's not necessary to reach medium or high vacuum.
This set up was for spheres of only 1 foot diameter and I don't think there are bags large enough for 6 foot spheres. However, my plan was to use a heat gun to stitch a bunch of the bags together to make it work. It's dirty but once again it should work theoretically. I was also planning on using a heat gun to section off portions of the bag to seal it around the sphere and cut off excess material but that part is really only necessary if you are about to achieve lift. I imagine it's possible once you've proven you can make a structure strong enough and light enough for lift that a better technique would be to incorporate a valve and find a way to dip coat the sphere to seal it. I never got this far.

A Potential New Approach To Foam

I mentioned experimenting with making foams and identifying polycarbonate as good material to turn into a nano foam. I use the term nano foam because aerogel wouldn't be technically correct. They are both nano foams. The aerogel is made using gel. This approach doesn't. It's very low tech and dirty. I theorized I could use the fact that polycarbonate is a thermoplastic to my advantage and mix it as a powder with another material that can withstand it's glass transition temperature but is also easily soluble in water. So, I found some polycarbonate powder (first American apparently to buy it) and mixed it with some ordinary table salt then put it in the oven. I know this sounds ridiculous. Then I washed the sample after it cooled in the sink and dried it with paper towels. Then I soaked it in rubbing alcohol and dried that with paper towels. Then I let it sit overnight to fully evaporate if it's a big sample. Then I weighed it. When I mix the powder in a 1:1 ratio by weight the sample after washing it weights exactly half of when I started without losing any volume. So I washed out all of the salt. But, that's not all. Because this method is basically sintering the particles together, it already had lots of air pockets in it to begin with. I attempted to make a one cubic inch sample to measure the density and it's not the most precise but the density is roughly 4.7 g/in3 which is about a quarter of the density of bulk polycarbonate. This means it's porosity is about 75%. It's not he 90-99.99% of commercial aerogel, but I personally find the initial results surprising. There's a lot of ideas I have to tweak this including playing with the mix ratio, grain size, uniformity of the particles, and aerating the powder. What I find very interesting about this technique in general is that it actually would work with anything that can be sintered including other thermoplastics, ceramics, glasses and metals. This means this approach could be used to make porous metals or even metal nano foams.
​

The 2009 analysis of the metal sphere UFO

I've recently been made aware of the 1994 spherical UFO that Steve Colbern published a report on in 2009. A few things stand out to me as someone who has been actively working on vacuum balloons and ways to make porous metals. First, it looks like two hemispheres nested inside each other exactly as I describe was my best approach to making a vacuum balloon based off of experimental results. Second, the sphere is presumably hollow. Third, the report clearly states that the sample analyzed was a porous metal with nanostructures present. A hollow porous shell with nested hemispheres of opposing orientation is exactly what I would expect a vacuum balloon to look like. There are ways to use my technique on titanium to make it porous although I haven't done so experimentally because it's melting point is very high. Materials other than salt could be used but even if salt was used it would be interesting because it would vaporize at the glass transition temp of titanium which actually might help make it more porous. I do believe Na and Cl impurities were present in the sample according to the report. Perhaps one could experimentally recreate this sample using this method (minus the isotopes.)
​

Crowdsourcing

If anybody wants to crowdsource the work on this with me I'm open to it. Also, if people are open to crowdfunding the research I'm open to that as well. Either way, it's up on the internet now. Maybe 10 years from now somebody as crazy as me will pick up where I left off. I might return to this at a later date, but without help I think I need to take a break.

I've been doing online research as well as some tinkering and was planning on building a prototype to demonstrate the first ever vacuum balloon, but I'm running into issues with expenses and time. I believe I've identified 2 approaches using well known materials that should work but one in particular that could be pulled off by a garage tinkerer with extra time and money to spare on the project.
Along the way I also started experimenting with creating foams using a technique I've basically invented as far as I can tell. I can't find any literature on it. I've gotten mixed results with it and am just not sure if it will ever work at least without being done properly in a lab setting. The approach has a lot of promise and I'll explain why.
There's a lot to go into on this subject. I've written about vacuum balloons before so if this is a new concept for you, you should give it a read.
I'm human so some of this work could have errors in it, but I have done experiments to test my theory and gotten interesting results. I have measured weight reduction in some of my designs and I have accurately predicted the results in cases where I could measure properly. That gave me a lot of hope to continue on at first but it's just a lot of work and I went way over budget early on. I can't keep pouring money into the project anymore and it hurts to say that because some of the results are so interesting. Also, life gets's busy and I can only tinker for so long.

Shapes

The best shape is a sphere because you need to withstand the atmospheric pressure outside the balloon pushing in at about 14 psi. For the same reasons we build bridges with arches, the sphere is the best shape for this because it will spread the forces out evenly. It becomes a matter of having a material that can withstand the compressive forces and in the case of non-uniformity (which to some degree is always going to be present) shear forces. Of course, the material also needs to be lightweight or it will never lift. Many sources will erroneously tell you no such material exists, but this isn't true. In theory, there are multiple materials that would probably work but the issue starts to become the total size of the balloon (and defects.) You could make it out of glass, but the balloon would have to be incredibly large and would be insanely prone to shattering and that's even if it was made defect free so there's really no point in trying normal glass. This is where choosing your materials is key so that you don't waste your time.
The volume of a sphere is V = 4/3πr^3
To calculate the buoyant force of lift at atmosphere you can simply multiply the volume by 1.29 kg/m3 and that will give you the amount it can lift in kg. Simply multiply by 2.2 for conversion to get the number in pounds. This formula was derived from the formula below.
https://preview.redd.it/6yf88k6uth4b1.png?width=516&format=png&auto=webp&s=0b5903bc3d27d74cc56765bcbe624c562d10cbab
The 1.29 kg/m3 is the fluid density of atmosphere and I simply removed the acceleration of gravity to show the force in units of pure weight rather than in Newtons. It's a simple calculation and understanding it is key to helping you design the vacuum balloon.
Now that you understand how to calculate the lifting force of vacuum in a sphere you can run a bunch of numbers and see for yourself that the lifting force is very small below radius 1 and grows exponentially above radius 1. This means it will be exceptionally hard to build a working vacuum balloon below radius 1 but unfortunately there are limitations to building large structures as well. Usually you want a prototype to be simple and cheap, not experimental in and of itself. This means the first demonstrated vacuum balloon will likely be about 2 meters in diameter or about 6 feet. It also means a vacuum balloon of very large proportions would potentially have incredible lifting force.
Now that you understand the relations between size and lifting force all you need to do is calculate the volume of the envelope of the spherical balloon. This is done by simply calculating the volume of a sphere of the size of the envelope and then subtracting that by the volume of the inner void. The difference is the volume of your envelope and you can easily calculate the weight of your envelope by multiplying the density by the volume. If you do this while calculating the lifting force and plug different numbers in you can easily see how the ratio of weight to volume works. You can also see how the density influences this and even can compare the volume of different shapes if you really want to just to see how much better a sphere really is than perhaps a square.
It's very important to point out that one of my biggest lessons in building prototypes is that there can't be any defects. I originally was making hemispheres and trying to join them together before pumping down to vacuum and every time there was a failure it was at the meeting of the two hemispheres. One solid piece seems to be necessary. It's conceivable that two hemispheres can be joined and bonded to become one solid piece free of defects, but I unfortunately did not have the materials to do this. I did do some experiments and found that you can reinforce this area with lightweight bamboo if necessary. However, these were small preliminary designs and I'm not confident that would scale well.
It's worth noting that the next best shape is a cylinder with hemispheres on each end. Basically a tic tac shape. It's only worth attempting this shape if you have reasons to from a manufacturing perspective. For example, I played around with the idea of making a foam sheet and then rolling it into a cylinder before it set rather than attempting to cast a foam hemisphere. It only makes sense if you are attempting a volume too large to pull off as a sphere for practical reasons (like it would't fit in garage or won't caste evenly.) Because it still needs hemispheres it's a design best left for after demonstrating a spherical design.

Materials

I dive into the use of aerogels and xerogels in the article referenced above. The purpose of these foam materials is because when engineered properly they retain a lot of their strength but lose a lot of their weight which actually increases their strength to weight ratio and that's exactly what we need to make this work. There is no material in bulk form worth pursuing for this design. You absolutely have to use a foam material. Even if you could pull it off using glass or beryllium, it's just not practical even for demonstration purposes. During my search I found the most attractive material in the bulk to be polycarbonate. It's still not worth trying in bulk form, so I invented a way to make a foam out of it. Polycarbonate is lighter and stronger than glass. Nobody has ever made an aerogel out of it that I'm aware of. I did not image my foam because I'm not doing this work in a sophisticated lab, but I can say fairly confidently that it's about 75% porosity. That's impressive, but I suspect that a lot of the bonding is weak and there's defects, but in my defense I used an insanely primitive and low tech technique.
There are two well known foams we all have access to that in theory should work. Styrofoam and polyurethane.
I understand that may cause you to sigh in disbelief. After all, polyurethane was invented in the 1930's at IG Farben and styrofoam in the 1940's so they are not only old but very ubiquitous. I should also point out that aerogel was invented in the 1930's and was once mass produced by Monsanto. None of these materials are new.
I used the given compressive and shear strengths published by a local styrofoam manufacturer to identify some common commercial grade foams that are very light weight that should work in theory if there's no defects. I tried working with them to have some custom shapes made, but they unfortunately are limited to 4 feet for one of the dimensions of their die blocks. This is very problematic even if we knew how to fuse two styrofoam hemispheres together. I'm not going to say it's impossible, but it makes pulling it off more challenging. I did do some experiments with small 1 foot diameter styrofoam hemispheres that are commonly available and managed to measure a weight reduction before it imploded. Anybody can replicate these experiments. I expected it to fail because the thickness was less than 1 inch. I found the best design was to nest two of these styrofoam spheres within each other but with the orientations opposing so that the point of failure for the outer sphere was across the strongest points of the inner sphere. This should create a perpendicular crossing of the hemispheres of the inner and outer shells. This is also where I tried some glues. Gorilla glue works best and sure enough it's a polyurethane. I was so impressed by it that I switched over to attempting polyurethane designs for the sphere.
I found a polyurethane foam used in boating that is only 2lb/ft3 which is very impressive. It also boasts a compressive strength of 38 psi. I figure that means half an inch of this stuff would be able to handle 19 psi theoretically. That's 5 psi above the 14 psi we need for our vacuum balloon. It's not a lot of room for error, but it works in theory.
What I like about polyurethane is that you can fairly easily make custom shapes with it and DIY. I experimented with a few different techniques and can say that you need this foam to be open to the air to set properly, but it does take on conformal shapes fairly well. The best method I found to make a hemisphere out of it was to actually blow up a rubber balloon and fit that snug into a styrofoam sheet for support and then pour the polyurethane foam onto it and let it set. You can then use cutting tools to clean up the extra material. This method works, but the cutting is a pain as I did it by hand. Precision will likely be necessary to properly join the two hemispheres and I learned this the hard way when I tried to join them. A more precise way to form the hemispheres I found was to buy plastic hemispheres and coat them in wax (to make removal of the polyurethane easier.) This is far more expensive than the balloon but gives more precise results. You can find people selling these in sizes up to 6 feet but it will get pricey. It's worth mentioning that I had a hard time removing the set polyurethane from the plastic even with a wax coating (which I also verified experimentally is the least sticky thing to use) so I'm not sure it's even the best approach. I've tried reaching out to polyurethane component manufacturers but so far no response. I'm sure outsourcing this would remove a lot of headaches, but also be very expensive for such a custom piece.
Just to highlight why I think this commonly available polyurethane foam is promising I want to calculate a 1 meter radius sphere of one half inch thickness to show that it should work in theory. Of course, this means no defects including the joining of the two hemispheres which is still a problem to solve but it's possible gorilla glue and precision would solve it. Maybe a DIY'er with their own CNC may want to give it a shot.
Using the volume of sphere formula given above we see that the volume of 1 meter radius is 4.187m3. The volume of a sphere of 1 meter minus 1/2 inch is 4.0295 m3. The buoyant lift of that is 11.44 lbs. The difference in volume (to find the volume of the polyurethane used) is .1575 m3 or 5.56 ft3. At a density of 2 lbs/ft3 that gives a weight of 11 lbs of polyurethane. That's less than the 11.44 lbs of lift.
I know what you're probably thinking. How does it hold vacuum? It's true that polyurethane and styrofoam are not expected to hold vacuum (I actually did find experimentally that styrofoam does hold partial vacuum for a few hours after it's shrunk much like the LANL aerogel) but you can simply wrap the sphere in plastic to hold vacuum. I planned on experimenting with dip coatings, but for experimental purposes I came up with a very clever design that I will explain later. Just know that the plastic doesn't have to be very thick to hold vacuum so it's very much within the range of possibility to coat the sphere in a thin plastic layer at less than .44 lbs. Plastic is very dense, but we are talking about literally a few mils of material. This is also why I roll my eyes at people who mock me for attempting a design with materials that don't hold vacuum. You are not limited to materials that hold vacuum for your design when you can simply add a layer for that later.

Experimental Set Up

I initially bought one of those vacuum chambers made out of a large steel pan and thick acrylic. Mechanical pumps are easy to find and relatively cheap. Mine came with the chamber. However, I quickly found it wasn't big enough and attempting to build a larger one looked costly. This is where I got clever and shocked myself with a very cheap set up that actually works. I simply bought regular large sized vacuum bags designed for storing cloths because they have a clever little self sealing mechanism that traps the vacuum. These bags are not meant for actual vacuum with a mechanical pump so I wasn't sure how it would work. I also had to find a way to rig it all up. As funny as it sounds my solution was to take the nozzle of an empty plastic bottle that happened to fit onto the hose and then I cut a piece of EDPM rubber to cover the end meant for the bottle and put a small slit in the center for air to move through. I then pushed this into the self sealing part of the vacuum bag and it actually creates a seal and pumps down! And when you remove the pump it self seals!
I found I sometimes had issues with pumping down properly and solved this by using a metal straw that I placed inside the bag near the seal and directed towards the sphere to act as a channel. Once again, to my surprise this works very well.
So, I then disassembled my original steel pot vacuum chamber and used the parts along with some parts I had to buy online to rig the pressure gauge into the system so that I could verify how much vacuum I was achieving. I'm a bit proud of this DIY set up because it works so well.
In order to properly record your results you must weight the vacuum bag and the metal straw as well as your experimental sphere before vacuuming. Then vacuum it down and pay attention to the gauge. If your design is not very good it may implode before achieving full vacuum. That's okay. You can actually measure a weight reduction without reaching the full vacuum. "Full" vacuum in this case is actually what is known as low vacuum. Low vacuum is all you need for a vacuum balloon to work as you have effectively removed most of the air and it's not necessary to reach medium or high vacuum.
This set up was for spheres of only 1 foot diameter and I don't think there are bags large enough for 6 foot spheres. However, my plan was to use a heat gun to stitch a bunch of the bags together to make it work. It's dirty but once again it should work theoretically. I was also planning on using a heat gun to section off portions of the bag to seal it around the sphere and cut off excess material but that part is really only necessary if you are about to achieve lift. I imagine it's possible once you've proven you can make a structure strong enough and light enough for lift that a better technique would be to incorporate a valve and find a way to dip coat the sphere to seal it. I never got this far.

A Potential New Approach To Foam

I mentioned experimenting with making foams and identifying polycarbonate as good material to turn into a nano foam. I use the term nano foam because aerogel wouldn't be technically correct. They are both nano foams. The aerogel is made using gel. This approach doesn't. It's very low tech and dirty. I theorized I could use the fact that polycarbonate is a thermoplastic to my advantage and mix it as a powder with another material that can withstand it's glass transition temperature but is also easily soluble in water. So, I found some polycarbonate powder (first American apparently to buy it) and mixed it with some ordinary table salt then put it in the oven. I know this sounds ridiculous. Then I washed the sample after it cooled in the sink and dried it with paper towels. Then I soaked it in rubbing alcohol and dried that with paper towels. Then I let it sit overnight to fully evaporate if it's a big sample. Then I weighed it. When I mix the powder in a 1:1 ratio by weight the sample after washing it weights exactly half of when I started without losing any volume. So I washed out all of the salt. But, that's not all. Because this method is basically sintering the particles together, it already had lots of air pockets in it to begin with. I attempted to make a one cubic inch sample to measure the density and it's not the most precise but the density is roughly 4.7 g/in3 which is about a quarter of the density of bulk polycarbonate. This means it's porosity is about 75%. It's not he 90-99.99% of commercial aerogel, but I personally find the initial results surprising. There's a lot of ideas I have to tweak this including playing with the mix ratio, grain size, uniformity of the particles, and aerating the powder. What I find very interesting about this technique in general is that it actually would work with anything that can be sintered including other thermoplastics, ceramics, glasses and metals. This means this approach could be used to make porous metals or even metal nano foams.
​

The 2009 analysis of the metal sphere UFO

I've recently been made aware of the 1994 spherical UFO that Steve Colbern published a report on in 2009. A few things stand out to me as someone who has been actively working on vacuum balloons and ways to make porous metals. First, it looks like two hemispheres nested inside each other exactly as I describe was my best approach to making a vacuum balloon based off of experimental results. Second, the sphere is presumably hollow. Third, the report clearly states that the sample analyzed was a porous metal with nanostructures present. A hollow porous shell with nested hemispheres of opposing orientation is exactly what I would expect a vacuum balloon to look like. There are ways to use my technique on titanium to make it porous although I haven't done so experimentally because it's melting point is very high. Materials other than salt could be used but even if salt was used it would be interesting because it would vaporize at the glass transition temp of titanium which actually might help make it more porous. I do believe Na and Cl impurities were present in the sample according to the report. Perhaps one could experimentally recreate this sample using this method (minus the isotopes.)
​

Crowdsourcing

If anybody wants to crowdsource the work on this with me I'm open to it. Also, if people are open to crowdfunding the research I'm open to that as well. Either way, it's up on the internet now. Maybe 10 years from now somebody as crazy as me will pick up where I left off. I might return to this at a later date, but without help I think I need to take a break.