Hacker News Reader

Comment by 6DM | Hacker News Reader

Breaking after a mere 10 miles, to me sounded really awful, but apparently they planned on this. The construction must have been to save on weight, making the wheels no more reliable than the exact distance they wanted to travel: "...milled out of solid aluminum. The wheels contact ground with a skin that's about half as thick as a U.S. dime, except at thicker treads."

View on HN

"[...] a mere 10 miles"

You must be kidding... this is a robot, which survived a hard landing, driven on exotic alien terrain, with async remote control from another planet :-)

The fact that they built it and drove it for over 15km is absolutely impressive by my engineering standards.

I think he's just commenting on the wheels, not the mission as a whole. We obviously are good at making wheels that last longer than 10 miles here on Earth.

Agreed but I would call the landing anything but hard. The precision and care with which that vehicle was delivered to the surface is just a mind boggling feature of engineering.

For anyone who missed this, they lowered the rover to the surface from a levitating rocket-powered skycrane.

I don't know if there's another true sentence from which I can derive as much joy as this one.

...which was controlled autonomously to boot!

Definitely one of the most bad-assed ways to deliver a rover to the surface of another planet. If you had told me as a kid this would be possible, I would have considered it almost impossible, except maybe as a science fiction concept.

Is the terrain any more dangerous or grueling than a desert terrain on earth though?

They drive it over desert-like terrains now to avoid such problems, but AFAIK the very first hole punched to one of the wheels was due to a small pointy rock. Martians rocks suck.

From my laymen perspective I'd say the terrain itself is less of a problem, the change in temperatures is much more extreme. This article from 2014 has many insights into prior Curiosity wheel damage:

http://www.planetary.org/blogs/emily-lakdawalla/2014/0819063...

This is a great article. Thanks for the link.

One thing it points out is that the rocks in that particular part of Mars are special: they are sharp and cemented into the terrain. There are places on Earth that have this kind of rock, but they are unusual. The wheels do fine in a rocks-in-sand scenario.

The article goes into a lot of detail, and it turns out that some of the obvious armchair science explanations for what is happening are not the case. For instance, the main constraint on the mass of the wheels was not keeping the liftoff mass of the payload under some limit.

Sharper rocks than a lot of Earth deserts - less wind and no water to smooth things down.

More broadly, though, there are some deserts on Earth that are pretty hair-raising. The NPS guidelines for driving deep in the Canyonlands (on roads, no less) warns that with a high-clearance 4WD pickup, you should still expect "considerable risk of vehicle damage". Curiosity is going a lot slower, but it doesn't have the luxury of roads or anyone to toss on a spare tire.

Depends on which trail in Canyonlands. Some of those are extremely technical crawl trails where you need a spotter and are planning every ten feet of movement, not unlike the Rubicon. White Rim is more of what I would call a road; watching a couple videos of people running it, I'd be less worried about damage there. (It's on my list for this spring.)

Honestly Jeep owners read a warning like that and get excited. Every scar in my paint is a new story, but to your point, getting the tools out in the backcountry is usually not fun.

Yeah, fair. I was specifically quoting the warnings for the Maze, which is remote enough that they desperately want to avoid tows and rescues.

I'd love to do White Rim whenever I get a chance, the overlooks onto it make it look wonderful. The Maze stuff looks much less friendly, but it's on the 'someday' list.

Oh, sure. The Maze was the area I was thinking of as well. I have a few thousand dollars of work and a lot more practice before I could comfortably consider those trails. I know my limits and I'd rather not hike out of Utah.

This is not obvious from the perspective of someone who's not familiar with rocketry or how light space structures need to be.

I'm curious(and this is in no way meant to belittle the work they've done) - if you drove a normal Earth car on the surface of Mars, how would the wheels and tyres behave in Martian conditions? Obviously we have off-road tyres here on Earth which can withstand some extreme abuse, so I'm just wondering - is Martian soil just full of sharp objects? Or is something else going on here?

Well, regardless of how well an Earth car would perform, it'd likely be far too heavy/expensive to get to Mars in the first place. It's impressive how NASA can solve problems with the resource constraints they have.

It would obviously be a dumb idea, but Curiosity is in the same ballpark as a car in both size and weight (1,982 lbs). A Smart Car is ~1,800 pounds, so it's actually heavier than some Earth cars!

Interesting, I didn't know that Curiosity was that heavy! It's a ways away from Jeep Wrangler/military Humvee territory (especially if you loaded one of those up with the required scientific instruments) but it'd be cool to see NASA go with something heavier-duty for longer-range missions in the future.

I think low temperatures alone would wreck Earth tires (and the rest of the car, but that's a different story).

It does have sharp objects, but ordinary truck balloon tires would handle those relatively well. But at night, the temperatures drop to -100 °C, and your rubber tires freeze and shatter.

Also, rubber's much heavier, and they just couldn't bring that weight. They really couldn't drop that weight from the sky crane, not without increasing the shock absorbers in the legs... which would have meant more weight, and the plan's just impossible.

It also has horrifically abrasive dust, much worse than anyplace on earth. The abrasion from that would wreck a balloon tire seal---hard to predict how quickly.

A balloon tire also goes from "fine" to "useless" with one puncture. Apparently they got surprised by the sharp rocks: http://www.planetary.org/blogs/emily-lakdawalla/2014/0819063... , not having seen anything like that from previous missions.

That is the best solution for Mars, but you might need to build the rover in a freezer. Tires that would be soft at Mars temperatures would be liquid or worse at Earth temperatures. For example, you might fill a normal tire mold with diesel, then cool it with liquid nitrogen and there you go -- a lovely tire for Mars that will melt at normal Earth temperatures.

You'd still want to go airless probably, due to a lack of repair shops.

Well, the tyre will almost instantly go soft as the inflating gas reduces in volume due to the cold. Then the tyre will go flat as the gas escapes (try driving for a couple of years without pumping your car tyres up). And the rubber will freeze and crack, and get UV perishment and low-pressure outgassing of its own.

The balloon-like tyre of choice is a springy basket mesh, as seen on the Lunar rover. But that's only needed because we wanted to drive around at human-acceptable speeds (to get to interesting places before the humans had to go home). With a teleoperated robot, we can take our time, and that means we don't need bouncy suspension. Which is good, because big wire tyres are quite heavy.

I thought NASA used metric units; would be nice if their press releases did too. I have no clue what "half as thick as a U.S. dime" means.

It's easily Google-able. A little more than half a millimeter.

Agreed. A number of millimetres would be a more helpful comparison.

It would be about .675 mm. A dime is about 1.35 mm thick

That's a description that most Americans can more easily picture than something like 0.68 mm (including people who are comfortable with metric units like myself) and they seem to be focusing a lot in recent years on making the information they provide more accessible to average Americans, undoubtedly to help drive interest in and support for what they're doing.

I'd prefer if they wrote something like "0.68 mm (about half as thick as a US dime)." That way the number is still there for people who don't want to look up how thick a dime is.

Good point, I should have added that translation in.

Are you implying you're the author of this press release? It looks like you were job hunting 6 months ago (based on your one HN submission) -- now you're writing press releases for NASA??

Edit: Nevermind, I'm a dummy.

There really was no need to go through the user's submission history. Try not to overreact next time.

I was interested to see his profile if he worked at NASA in this capacity. He had one submission -- it's not like I pored over his history. I don't think I overreacted -- I was impressed that someone could get to a point that they're authoring NASA press releases so quickly.

Like I said, though, I'm a dummy...

It's very likely that the mechanical drawings were in inches for the machinists, even for NASA. The scientists are using metric units.

No. That would not be the case.

What makes you say that? I've worked for two spacecraft companies in the US who supply parts (and whole spacecraft) to NASA. Both of them did drawings for small parts in mils (thous) for both electrical and mechanical parts. It seems like most machinists in the US are used to working in mils.

I remember on my first day (year 2012), I was shown an E-sized drawing of a spacecraft bus at 1/10 scale. The unit was inches. I was surprised it was in inches and said something. The person showing me just laughed and said welcome to aerospace.

The Wikipedia for mil is full of machinist lingo: https://en.wikipedia.org/wiki/Thousandth_of_an_inch

The mechanical specifications I've seen for newer (post-2000, say) components at the lab that built this particular robot have all been in metric units. I just checked, and the current requirement is that metric "shall" be used, unless (a) a particular mission waived that practice - I don't think MSL did; or (b) for parts from "industries that are historically in another unit of measure." Two examples I know of are gears and printed wiring boards, which are indeed in mils.

And you're right, it is possible, barely, that the wheels were an exception. I don't know if they were fabricated on lab or not.

Thanks for the explanation and glad to hear that space is slowly moving towards metric.

Maybe they should have a second, international press release.

Just a "10 miles (16km)" and "half as thick as U.S. dime (0.68mm)" - either directly in text or as on-hover/on-tap tooltips (for web) - would've did the trick just perfectly.

"half as thick as U.S. dime (0.68mm)" is ambiguous: is the dime that thick, or the wheel?

Or you could say something like:

...10 miles on a tire only 0.68mm thick, about half as thick as a US dime.

I'd give them a pass for the press release -- it was paid for with U.S. dimes.

Remember that the rover weighs almost 2000 pounds, so each wheel is responsible for carrying around 330 pounds on average over very rough terrain.

Is that the weight on Earth or Mars? Sites listing its specs are unclear on that.

Wikipedia says the dry mass is 899 kg, which is about the 2000 pounds of weight (on Earth) mentioned by GP.

https://en.wikipedia.org/wiki/Curiosity_(rover)

Which would make it about 338 kilos on Mars - wikipedia lists Mars' gravity as 0.376g (vs earth's 1)

Technically that would make it 899kg on Mars, the kilogram is a unit of mass. However the Rover would weigh ~8800N on Earth and ~3300N on Mars.

The pound is a unit of mass, same as the kg (it's defined as exactly 0.45359237kg). Just as 900kg is 900kg everywhere, 2000lb is 2000lb everywhere. The colloquial weight is literally a force, and its imperial unit is thus the pound-force.

This is incorrect. The pound is a unit of force, similar to a Newton. The pound-mass(notated lb-m) is similar to the kg. The conversation you gave above is based on Earth gravity for weight.

> This is incorrect. The pound is a unit of force, similar to a Newton. The pound-mass(notated lb-m) is similar to the kg.

That's the pound-force (lbf).

Pound and pound-mass are aliases. Indeed the official facts sheet defines the mass of the rover in kg and pounds.

> The conversation you gave above is based on Earth gravity for weight.

My comment is based on mass being an intrinsic property of matter, gravity does not enter the equation.

There are about a dozen different units that are all called "pounds," and you have to rely on context to know which is meant. When I took physics in high school, pounds were always force (abbreviated lbf) and we used slugs for mass ( https://en.m.wikipedia.org/wiki/Slug_(mass) )

It's the weight on Earth. Just look for a source that specifies in kg - there's no ambiguity about what the SI unit represents.

No, ambiguity was introduced when the measurement was switched from force to mass. The question remains: what is the weight/force experienced on the wheels on Mars? Knowing the mass won't tell us unless we also know the acceleration due to gravity on Mars.

Yes, but it looks pretty tore up and the treads seem like they're coming completely off. I was under the impression that the wheels would have been built a little more sturdy. I know it's not the same, but car tires last for thousands of miles. So you would think that a hyper efficient tire, battle tested on earth terrain, would be designed to last at least 100. This is after 10 miles.

Edit: To clarify, I know they are thin and to save on weight. I just never thought it'd be designed to only last 20ish miles. Though another comment has said it will run fine without a lot of this material.

It's been four years, the wear could be as much a function of time as of travel. Car tires on earth dry out, crack and disintegrate over time.

The skin is "about half as thick as a U.S. dime". A car tire that thin wouldn't last a 10 mile drive.

The tires are extremely thin to save on weight. Take a look at a dime and think half of that!

the wheel skin is 0.75mm thick, and was specifically designed to a goal of 10 to 20 km with the smallest weight possible: the rover mass is 899kg, the launch vehicle totals was 531000kg - by reversing the payload fraction one can see that for every gram of rover one needs more than half kilo of rocket to get it to mars, so there's that - you can get outside of your launch weight real fast.

this article has all the answers you're looking for:

http://www.planetary.org/blogs/emily-lakdawalla/2014/0819063...

short story - the wheel were tested to specification against expected rocks etc but once on the boogie suspension the front wheel had to support weight plus the force of the other wheel pushing, so there's where the extra punctures come from, this is what force was not accounted correctly.

Thanks for the reminder of scale. Without a reference, it's easy to forget how large this rover is, in pictures it looks very cute.

Here's a good reminder:

  http://i.imgur.com/3Hg9O.jpg

This shows engineering test models of Pathfinder, MER (Spirit/Opportunity), and Curiosity. It was taken in the Mars Yard at JPL, which engineered all three.

I wonder if stresses build up from the day night cycles on Mars - which I suspect would be pretty harsh on finely machined bits of metal.

Maybe that's why it's a metal and not something more traditional.

and lots of radiation

Sort of. The early wear was much faster than anticipated, so they've since been using softer ground (sand instead of rock) as much as possible. You don't want to carry around any unnecessary grams on a rover, but I suspect the next one will have sturdier wheels.

I'm sure you would have done a better job. Now please tell me: How would you have improved on this given the mission constraints?

Are you 100% sure that this isn't a case of 'why didn't they?' where laypeople not involved in a project know exactly what should have been done by the experts that were involved in the project?

Why the harsh tone? The parent poster outlines their initial incredulity and subsequent learning. Nowhere do they suggest that they could have done better.

Let's also note that the top speed of the rover in optimal conditions was 0.09 mph (0.14 km/h), and it took them a full year just to reach the first kilometer of travel.

They planned on this, but the article (and NASA) does not reveal whether the wheel damage was expected at this point or sooner or perhaps even somewhat later.

It was much worse than initially expected.

Reply on news.ycombinator.com