Because I think that the engineering problems facing hyperloop can all be overcome, I’m going to discuss several criticisms I found on the web and ways they can be overcome. I’m only going to look at serious criticisms and not hand-wavy “this couldn’t possibly work!” criticisms that can’t be answered anyways.

After scanning a couple articles on the topic, they all seem to be from 2016 shortly after the white-paper was released and nothing of substance more recent. If you know of such an article, please email me with a link.

I’ll start with the criticisms in this article, which helpfully has a link to the original post by a Kristen Ray:

Thermal Expansion

Thermal Expansion: The paper claims that no expansion joints are needed because “tube is not rigidly fixed at any point” and uses “dampers” instead? A 20 degree change in temperature would cause an expansion of 600 FEET across the whole length of the steel tube. A “telescoping tube” like the ones they use for airplane access is not a good fix for those sorts of distances and tolerances. And a damper attachment to each pylon would not allow for that much movement.

The first criticism is about thermal expansion and I think it is a valid concern, but straightforward to overcome. I have two ideas: Kevlar-reinforced vacuum-rated rubber expansion joints and radial gridiron joints. The first is pretty much self explanatory, so I’ll only talk about the second one.

The gridiron pendulum was invented around 1726 and was used in precision clocks for maintaining a (nearly) fixed length pendulum even in the presence of temperature changes. It did this by using two dissimilar metal rods arranged so that the expansions cancel out. It should be possible to construct a similar structure that is completely sealed and sufficient for the purposes of counteracting thermal expansion in hyperloop tubes.

Of the two, the Kevlar/rubber joint is probably going to be both cheaper to manufacture and easier (another way of saying cheaper) to install.

Seismic Activity

Seismic Engineering: Their “structural simulations” (figure 15-20) are clearly not to scale for the average span (between 20-100ft tall and average of 100ft between spans). A simple damper will not suffice for the major seismic activity potential in the San Francisco and Los Angeles areas with a structure that needs to keep such exact tolerances due to the high speeds. Most likely this will need to be engineered for this specific system, and it won’t be able to just pull an exact solution from other structures.

The original white-paper and subsequent developments have the pod being either floating on an air cushion or magnetic levitation. Both of these will themselves dampen seismic activity’s effect on the passengers. It should be possible to design some additional dampening into the support structure to dampen the initial seismic activity. Add to that a system like the Japanese high-speed Shinkansen that brakes when an earthquake is detected. There are structures both taller and longer than the pylons discussed in the original white-paper that withstand earthquakes, so this is a solvable problem.

Cost

Dampers: The Dumbarton Bridge in SF Bay Area recently had 96 bearings installed under the road deck for a seismic retrofit. They cost $90,000 EACH [1]. If one of these bearings was on each pylon, it would add over $2 billion. And there would probably be two bearings per pylon (one for each tube), so we’re probably at $4 billion.

Note: the link [1] from the original page is dead.

This is referencing a government project, and anyone familiar with government projects knows this number is massively larger than the actual cost of the materials. These bearings are probably massively over-engineered and multiples of the manufacturing costs to cover things like contract compliance, busy-work jobs creation and political kickbacks. The bearings in question if designed without government involved would probably have cost between $5,000 and $15,000 a piece. If a hyperloop is built with minimal government interference, this is the kind of cost savings to expect.

Concrete Pylons: He’s averaging $100,000 per pylon. I don’t have any numbers to compare this to, but that seems very low to me.

Given that I am not a Civil Engineer and don’t routinely work with concrete, this doesn’t sound too low of a price to me. Concrete is nearly as cheap as dirt and forming the pylons are a matter of building the re-bar, setting up forms, pouring the concrete, letting it set and then tearing down the forms to repeat with the next pylon and ideally doing several instances of this in parallel so it doesn’t take forever. All this is done with overpasses in the interstate highway system now, so there is nothing fancy being done here.

Communications

System Communications and Power: Most rail lines have electrical and fiber lines running along the track to power the system and to have a real time map of everything happening along the line. The preliminary designs call for a WiFi communication system, but that is not as reliable as fiber or microwave communications. It will also need substations, communication centers, etc.

WiFi operates at either 2.4GHz or 5GHz, both of which are microwave frequencies. In fact, your microwave oven operates in the 2.4GHz band that your WiFi router does. Being a sealed metal tube makes the hyperloop tubes wave-guides and will allow existing WiFi equipment to be used at much longer distances while also isolating it from interference from outside the tube. In addition, running fiber alongside the tubes will be reasonably cheap. What I would expect is this: use commercial WiFi transceivers with a firmware change to allow them to operate at frequencies just outside the ISM bands they normally operate in for communications between the pods and regularly-spaced base stations. These base stations are then linked with fiber to the system control centers.

I don’t think that the original white-paper’s idea of onboard batteries powering all the systems on the pod is a good idea. I think it would work fine for short distances of a few hundred miles, but not for longer distances. Instead, I think either a high-speed pantograph, or more likely, an inductive power transfer system will be needed if hyperloop is able to become a nation-wide network where you can board a pod in New York City and step out in San Francisco.

More Criticisms

In the same Quora thread, we have Sandeepan Majumdar with these criticisms:

Vibration

Vibrations: When you are moving at Avg speeds of 900 km/hr the machinery mounted in front of the capsule has to be most awesomely balanced machine ever made. Not a mechanical engineer so won’t comment on that.

Both supersonic aircraft and rockets routinely travel at these types of speeds and they don’t tear themselves apart doing so baring rare catastrophic failures like bird strikes or O-ring failures at low temperatures.

Cost

Economics: I have read the dumbed down version in the white paper. Here is an equally dumbed down rebuttal. We have built a long vacuum chamber before. Its called the large hadron collider. Its about 27 kms long and cost 6.5 billion. Now I know economy of scale argument is legit. But when you scale up the LHC to hyperloop size and add additional complexities such as security and thermal expansion, vibration. You would make this an incredibly complex operation. That just means cost needs to be added on the existing project.

Quite the opposite, actually. Hyperloop will be less, not more, complex than the most complex high-energy particle collider created by man. The cost of the Large Hadron Collider (LHC) includes more than just a vacuum tube. It is specialized for high-energy physics to contain particles traveling near the speed of light with massive superconductors for magnetic confinement at vacuum levels much deeper than what the hyperloop would use. For 350 miles (563 km) of steel tubing and standard commercial-grade vacuum pumps, Musk still was estimating (an educated guess) several billion.

Yet More Criticisms

The Daily Caller has this article about chemist Dr. Phil Mason’s criticisms, excluding the repeats of the thermal expansion and cost concerns that are already listed above:

Everybody Dies!

Musk’s Hyperloop project “might be better described as all the problems of space travel while traveling in a gun barrel at the speed of sound,” Dr. Phil Mason states in the video. “Any failure whatsoever will rip though that 2 centimeter outer tube like candy. Now sure, anybody in the capsule would die pretty much instantly in the event of a crash…but a single breach in the Hyperloop would probably kill everybody else in the Hyperloop because air would rush into the tube at about the speed of sound.”

I don’t have an answer for this, but instead a series of questions:

“Any failure” seems overly broad. What set of failures do you foresee as being an issue? I can think of the following:

• Rapid depressurization
• Pod crash/pile-up
• Loss of power/communications
• Passenger medical emergency

Has any of these been tested? Would an emergency system that both brakes all the pods and releases the vacuum mitigate this risk?

Merely shooting a few holes in the thin tubing surrounding the Hyperloop’s vacuum would create air pockets which would trigger the same kind of cascading failure caused by a crash.

Incredibly tiny holes created by modest rifle grade weaponry could trigger the kind of cascading failure that would kill everybody in the system. To make matters worse, the 373 mile length of the Hyperloop and the fact that it would run down the middle of the freeway would make it effectively impossible to defend from terrorists.

Dr. Mason compared the terrorism risks of the Hyperloop to air travel by saying “any crazy with an anti-material rifle could shoot holes in the tube which would probably be fatal to almost the whole system.” In comparison, Mason noted that “one plane crash does not destroy the entire infrastructure and kill everyone else flying the same route.”

Any large object will have a terrorism risk, but the actual effect on the system will be one of the above failure modes, most likely rapid depressurization and lost of power/communications. Mitigate these and you’ve mitigated the majority of the destructive power of terrorism against hyperloop.

Regardless, the more you let fear dictate your actions, the more power you give to the terrorist, both tyrannical governments and the suicidal maniacs.

Can’t Do It!

Its Probably Physically Impossible To Build The Hyperloop

For the Hyperloop to work, it would need a way to pump out roughly 2 million cubic meters of air from its tubes and make sure that the air stays out of a 373 mile-long pipe with walls less than an inch thick. In comparison, the world’s largest vacuum chamber only pumps out about 1.5 percent as much air and requires enormous amounts of structural reinforcement.

The structural reinforcement required is proportional to the size of the chamber. The referenced NASA chamber has a 100ft diameter and is 122 ft tall. For that size, a 1 inch ring of the chamber has to support 55,417 lbs of force, whereas the 10 ft 10 in hyperloop tube has to support only 6003 lbs of force which is well withing the capability of an inch of steel. The length of the structure doesn’t matter, as another inch of tube has another, identical set of material to carry the load of the new force. This is why pressure vessels are almost always shaped like cylinders with rounded ends, the main exception being spheres.

Note: force ≈ 14.7 lbs/inch^2 * π * diameter (in inches) * 1 inch