Yet another fixed gantry machine...

[handlewanker]

Hi

I think my initial post and topic is missleading. Sorry for that.

To clarify, here is my goals in list-form:
- It is a hobby project
- For making small wood, plastic and aluminim parts.
- No steel milling intenden.
- Hobby machine on a hobby budget so i´m not willing to hire engineers etc
- Will try to design the build to use materials i allready have.

I will talk to my friend who is helping me with machinig and come back here with an updated sketch.

Hi, in the drawing you posted in post #1 I think your approach with having the bearing blocks on the base and linear rails on the table is unstable..........better to have a long solid base with the linear rails on it and the bearing blocks under the table....that is Hiwin type rails too, not the SBR round rail type .

That means the base is twice the table travel length......in your case you wanted a 400mm X 250mm ( 16" X 9") work envelope, so the machine base will be 800mm long with a 250mm + space between the column uprights......personally I'd go to a 300mm table width

The bearing blocks can be spaced apart 400mm overall on the log axis and slightly inboard from the table sides.

This will give you the table travel of 400mm on the 800mm rail length, very rigid with bags of support.

It's better to have the longest axis as the Y or table travel to reduce the width between the side support columns.

The table also needs to be high enough above the base to clear the ball nut housing, so this means 20mm Hiwin rails sitting on 20mm thick steel strips 40mm wide and 800mm long.

The table needs to be longer than the travel to cater for a 4th axis if it's required or else you will reduce the table working area significantly.........that means a table length of at least 400 + 200mm.....100mm + each end for the 4th axis body and the tail stock to give you the 400mm working area in the middle........I wouldn't go higher than 125mm max clearance under the bridge for the work you intend.

The table width at 250mm (300mm?) will also mean the columns needs to clear the table by at least 10 or 20mm each side.

Acceleration forces will be hampered by table weight, so if you have a table of 20mm thick steel plate that does not give you any depth to have tee slots.....going to 40mm thick and the table will weigh a ton.

I think an aluminium slab 40 mm thick would be a starting point, and my choice of clamping facilities would be to have tapped holes as opposed to Tee slots......the parts you intend to mill will not require tremendous force to hold them down.

In the design I'm working on I will have parallel holes with expanding type work holding bolts......this will ensure the holes can be kept clear of swarf etc as opposed to tapped holes that clog up..

All of the moving gantry routers on EBAY have tables formed from extrusions bolted together, it just depends on your preference, but you'll want to have a surface that is flat and can stand some wear and tear without distorting etc.

There's dozens of different ways to make an XYZ movement machine, but it needs to have good rigidity to do any milling.

I assume you are not restricted by space or weigh limitations........that is a bummer.
Ian.

[Hartzell]

Thank you very much for sharing your knowledge!

I have redesigned it a bit.



Milling the linear surfaces is an issue. What do you think of the following approach?
- Make the gantry and base as separate structures that’s bolted together
- Bolt it together before welding, then separate it for milling the linear bases




I wont be able to machine the bolted surfaces, is that an issue?
Can I mount the linear rails direct on the frame or is it necessary to have steel plates?
Do I have to heat threat the structure after welding to reduce stresses? (I don’t even know what the process calls in Swedish, but I hope you understand what I mean. )

[wizard]

Your pictures aren't coming through, at least not on my iPad this morning. So my comments are based on your text only.

Thank you very much for sharing your knowledge!

I have redesigned it a bit.


Milling the linear surfaces is an issue. What do you think of the following approach?
- Make the gantry and base as separate structures that’s bolted together
- Bolt it together before welding, then separate it for milling the linear bases

I suppose you could do that, I'm not sure what you would gain.

I wont be able to machine the bolted surfaces, is that an issue?

It can be but there are work arounds. One example would be to make the mating surfaces in such a way that they can be grouted once the machine is aligned. Here I'm thinking a form of epoxy as the grouting material. I would lean towards preferring that the mating surfaces are machined but there are other solutions that would lead to a properly aligned machine.

Can I mount the linear rails direct on the frame or is it necessary to have steel plates?

It depends. Steel plates are often used for a couple of different reasons. One is to provide enough to material for the bolts that hold the linear rails on. The other is to allow for a machineable surface that allows for a flattening of a structure. If you end up with a structure that doesn't distort the frame and can support the rails properly they the steel plates aren't needed.

The question is an interesting one because I've seem commercial machines built out of steel tubing that use two different methods. One axis used steel tubing with plates welded on and machined. Another axis used heavy steel tubing with the mounting surfaces for the linear rails machined into the tubing.

Do I have to heat threat the structure after welding to reduce stresses? (I don’t even know what the process calls in Swedish, but I hope you understand what I mean. )

That would be stress relieving and in an ideal world the answer would be yes. With care and the right materials you can get away with no stress relief. This is where bolted up designs can have real advantages.

[LeeWay]

Consider also that welding does not have to be complete full closed off welds to make a viable structure. Spot welds and skip welds work great on automobiles in certain areas. The same can be done on a machine. What this does is lessens the heat distortion. That lessens the need for heat treating to ease the stresses inside the steel. Just an option to consider.

[Hartzell]

OK. Maybe i should build and machine the gantry and base and then spot weld it together instead of bolting?

I also found some 125x65x5 tubing i could use to the gantry frame, a little beefier than 120x60x3 if thats doing any difference.

[LeeWay]

Not necessarily spot welding. Skip welding. I am not saying that is what you should do. Only suggestions for things to consider that may make it an easier less costly hobby build.

[handlewanker]

Hi, your latest design is looking much better from a construction angle.

One thing I'd avoid and that is cutting into the main sides of heavy tubing.....it takes away the linear strength of the tube .......the cut out at the bottom to the back is what I'm referring to.......cutting it out will be a job in itself.

I see you are going to bolt the bridge and uprights to the sides of the base......that means you also need to weld thick pads to the base sides to bolt against.

With the bridge and columns as a separate complete assembly you can machine the faces of the X axis rail seats easily, but it will be impossible to mill the inner faces of the column bolt faces, and machining them first and bolting them on and further welding is not recommended.

Also, to get the column bolt faces an exact fit to the base sides with a fully welded bridge/column assembly will be a task beyond the average machinist even, and moving the bridge/column assembly around to mill the other face will make the whole alignment and squareness a bit iffy.

I opted to bolt the columns on the top of the base which makes machining the various faces straight forward.

If you look at the polystyrene model I built in a previous post you'll see where the column base bolt faces occur........with this build I'm drilling holes through the bottom of the base section and passing cap screws through the upper face to bolt into the column bottom.

The top face of the base has 20mm thick steel pads welded to it to make mounting points for the columns......the column bottoms base plates being tapped for the bolts.

These are just a few of the pit falls you'll come upon if you don't design with the knowledge of how it goes together and is secured etc..........you are going to make it........design with actual making in mind.

I made the polystyrene model over a week end purely to see if the machining and assembly were a practical proposition, and also to see the range of the various slides etc.

I would suggest that you do a cardboard full size model as it's easy to cut and try and glue it together.....once you see the shape of the machine in the flesh you can position the table on the base section and see how it reacts with the bridge assembly......you'll also have to project a pseudo mill spindle on the bridge face to see where it positions to the table travel..........this will affect the column fore and aft position on the base.

Going into metal is the last step and is easy if you have a definite plan, but designing on the fly is not all that clear for the positions of the various components etc.

BTW, if you bolt the linear rails to the top of the side tubes this will mean the columns have to be higher, whereas if the rails are bolted to flat steel bars welded down inside onto the base cross members the columns are shorter and more rigid.

You'll have to ensure the side base tubes are spaced apart enough to clear the sides of the table by a minimum of 10mm, and also ensure the top of the table is slightly above the side of the base side tubes by at least 10mm., and allow for the ball nut height to clear the base cross tubes when it's bolted to the bottom of the table.
Ian.

[handlewanker]

Hi.....looking at post #22 and a coup[le of things could be made easier.

1st,,,, get rid of the cut out on the long tubes.
2nd.... get rid of the cross tube at the bottom of the gantry columns and the flat plates across the ends of the long base tubes.
3rd.... add another cross tube to the base in place of between the columns......the bridge and columns are now one piece in a large inverted U shape.
4th.... machine the faces of the pads on the base for the column bolt faces.....as squarely and parallel as possible
5th....make the columns first as two separate parts, machine the lower bolt faces and bolt them to the base with at least 6 bolts per side.
6th.... make the bridge and machine the linear rail faces and ball screw bearing support pads etc.
7Th...fit the bridge between the columns and align it for vertical and axial squareness to the Y axis, then tack weld it in place.......check alignment and progressively weld the bridge from side to side alternatively to the tops of the columns.
8Th....check the alignment of the bridge and column assembly to the base linear rails and adjust the columns on the bolt faces.....then insert dowels to the column bolt faces to keep the alignment square.
9th....forget about stress relieving as there will be no further cutting of the frame once it gets to this stage.

You will of course need to have at least 20mm thick linear rail seats (steel flat bar) on the face of the long base tubes and the face of the bridge.....welded with short 20mm welds at 50mm intervals along the length both sides.

Make the bridge flush with the face of the columns not set back to allow full travel across the face of the bridge.

The ball screw bearing mounts at each end are bolted to the base end cross tubes so allow for mounting pads on the two end tubes....same for the bridge.....add two pads for the end bearing supports etc....and machine them all in one setting.

Hope this helps.
Ian.

[A_Camera]

Hi!

I have given up my moving gantry plans and started all over.
I want a small, rigid machine for making panels in plastic and aluminium. Aircraft type of panels. And some wood decoration.
Work area around 16" x 9".
Accuracy is prioritized over speed.

First question:
What do you think about this construction?
http://www.romanblack.com/CNC/01_72.jpg

Hi Hartzell,

Been there done that...

I know exactly how you feel, I started also with some ideas about building a CNC with moving gantry and after trying out different solutions I have given up all together. Restarted the plans with a fixed gantry design in November last year and my CNC is basically ready and works fine today. My aim was about the same as yours, precision, accuracy and repeatability had priority before speed, work area about the same as yours 250mm x 350mm ( 10" x 14"), and for what it's worth, at least for now, it seems that my design worked out just fine. If you are interested in details you can visit this blog. The story of my CNC project is long, contains so far 5 parts and is not finished, there is going to be at least a part 6 also.

The first part is: Adapting Camera: My CNC story (part 1) and then it continues to the next and so on.

Part 6, apart from updated pic will contain a warning. In part 5 I speak warmly about a USB controller, which I truly liked from the start because it seemed to work nicely. Well, yes, it works well if you don't want to implement any auto tool setter feature. The G31 probing code is not working at all and even the built in Z height probing has SERIOUS bugs in it. I came up with a fix and solved that problem, even if my solution isn't as good as if the G31 would have been properly implemented. Another very serious issue, even more serious than everything else is, is the fact that the card can cause the CNC to run away totally uncontrolled. This can happen if you start a vacuum cleaner or switch on some other appliance connected to the same phase which has an inductive part in it, like something with a motor or transformer in it, or a florescent lamp or something similar. The results can be that the CNC movement stops on one or all axes or in some cases one or more axes starts using rapids. Uncontrolled movements are rare but after that they happened twice in real use and a few times while I milled air I decided to remove it and return back to ordinary parallel port use. I don't like that solution with parallel port because it is a dinosaur, but it is reliable and never caused any problems. The only problem is that Mach3 is not working with any laptops, at least not mine, which is a modern HP, even though I have a parallel port in the docking station which every other software is capable to use but not Mach3. I don't know why that is and not bothered to do anything about it right now, I am using an old desktop computer at the moment.

Anyway, my advice to you is to be very critical to what you read and which advise you chose to follow or discard. For milling plastic you don't need to make a complicated steel base, I am pretty sure that my simple design would work even for your needs. I aimed at milling plastic only (mainly POM or Delrin) but after I have used the CNC for a while, I am pretty sure it works well even for aluminium, though not tried yet. I would never get myself into welding, I went for aluminum frame and using of T-nuts and bolts/screws. My maximum milling speed is 450mm/min (about 18"/min) and for the rapids I can drive with up to 550mm/min (about 22"/min). This is more than enough for such small machine in my opinion. Have a look at my blog and if you are interested in more detailed description of some part, or specific details or pictures, let me know and I'll try to answer your questions. There is a lot of text, perhaps a bit too much, but like I said, the series is still not finished since I am now ready with the machine, and also abandoned the use of USB.

[A_Camera]

Thank you very much for sharing your knowledge!

I would not get into welding, but if you are very good at it then it may work out fine in the end. Never the less, if you want to reduce material and weight you could leave out that middle beam. The way I see it, it has no function since there will be no side forces there with a fixed gantry.

Edit: I think that the above is actually true for the other two beams also and I don't think they are necessary at all. There will of course be a need for something to hold the lead screw, and there will be some pulling forces on the end plates, so those must be rigid enough, but for milling plastic and aluminum I think that you could save quite some weight. Remember that in end, you will need to move this as well, so start already now to calculate weight and means of moving it around.

[A_Camera]

Hi, so very true.......always design from the standpoint of how your going to make it.

The principle of any design must be established first, then the design will work, otherwise it's a castle in the air.
Ian.

So true, so true...

Not only that, but my experience is that if you think about making a DIY CNC you have to start with your own knowledge base, tools and machines you have and the capabilities around you. It is pointless to sit down and design something if in the end it will need tools and machines which costs a fortune to buy and you can't afford them, or if you lack the knowledge to use them properly. I learned this during my own project, and even though at start I was convinced that I had all the tools and knowledge I needed, I had to buy more and some new tools/machines because what I had was simply not good enough for the task even though I thought from the start that they will be. OK, if you have a fully equipped workshop with a lot of nice machines already, but if not... then you need a lot of good friends who have them, or money to buy and space to place them and to use those tools. This sounds really basic, but I think many people miss this kind of thinking and end up with that castle in the air. There are plenty examples of that on this forum and elsewhere.

[A_Camera]

OK. Maybe i should build and machine the gantry and base and then spot weld it together instead of bolting?

I also found some 125x65x5 tubing i could use to the gantry frame, a little beefier than 120x60x3 if thats doing any difference.

Why not skip welding all together? Sounds like you are just as much a welding expert as I am, and the last time I tried welding was about 40 years ago... that's how big of an expert I am.

Never the less, I don't understand all this about the need of welding in a hobby machine, unless you love welding and know how to do it properly and I misunderstood you. You mentioned that accuracy and precision is important for you. I think, if you not already have one, you should buy two or three major tools for your workshop, one good (but nothing extreme) bench or pillar drill, a good, but nothing extreme, tool to be able to cut the material you will be using, plus an angle grinder. Drill wholes, tap them and screw/bolt the pieces together. Drilled and bolted properly and you can build bridges to hold many tonnes, there is no need to weld and there is no need for 100% rigidity.

One HUGE advantage of bolting the pieces is that regardless of how careful and cautious you are, you WILL be making some mistakes. Imagine if you have welded everything together and you suddenly realize you missed something and one of the pieces are wrong or needed some small adjustments. What are you going to do? Start a new build or cut it up again? I am talking about less than a millimeter adjustments, which might be necessary to make to get the accuracy you need. Those inaccuracies will not show up before the end, and if you have welded everything together then you must have a plan to correct the CNC. Now, I may sound pessimistic, but my experience is that adjustments will be needed. Do you have a plan for that?

By screwing the pieces together the worst that can happen is that you need to replace one, or several pieces, but most probably, if you have drilled the holes right, you will be able to adjust using the screws you have used. Some inaccuracies for wood decorations is no big deal but if you will be making aircraft models, some some other things which demand accuracy then you have to start thinking about how to get that accuracy, not just on drawings but also in real life, so you really need to figure out a way of making fine adjustments when you start milling.

[handlewanker]

Hi, by choice I'd never bolt anything together as an assembly.....too many points to drill and tap and align etc.

Tapping into anything as thin as 5mm will limit the diam of the bolt you can use.

If it's going to get to the starting point we need a final drawing of how you want to make it, that will determine the design and also the feasibility of a bolted or welded structure.
Ian.

[A_Camera]

Hi, by choice I'd never bolt anything together as an assembly.....too many points to drill and tap and align etc.

It is true that the number of holes to drill, tap and align are many, but at the same time it does not only have disadvantages. It takes longer but the end result may be much better, or at least as good. The time which seems to be wasted at the beginning can be a HUGE gain at the end of a DIY CNC project. I am no expert in welding, but the intense heat most probably and definitely changes the shape of the beams. It is a difference between the theoretical precision and the accuracy Hartzell is talking about and I doubt you could get that accuracy in a welded frame without a lot of pain.

As for the number of alignments you have to make, well... I really don't see any difference in this matter because if you plan to weld with any accuracy then you have to align in advance and the alignments and the fixture of the alignments must be EXTREMELY accurate at all joints otherwise in the end you can't do anything else than cut the whole thing up again and restart from where it went wrong. Remember that some of of the inaccuracies and mistakes won't show up before you start milling, and once you are there it is pretty difficult to make a slight alignment change if everything is welded.

Tapping into anything as thin as 5mm will limit the diam of the bolt you can use.

Yes but you can ALWAYS use a nut or come up with a solution. Remember that we are talking about a small CNC. Actually not once during my own project I wished that I could weld instead of drill/tap/bolt. Yes, I have used hundreds of nuts also, but so what? Using nuts has the advantage that the drilling accuracy can be lower apart from cases where it is necessary to use them.

If it's going to get to the starting point we need a final drawing of how you want to make it, that will determine the design and also the feasibility of a bolted or welded structure.
Ian.

What I also learned is not to rush... I rushed and there aren't many parts in my final CNC which comes from the originally designed and made one. I learned by my mistakes and on a tight budget those mistakes would have been fatal and could have stopped the whole project. Drawing is one thing, doing is another. The drawing must start after the design and is supposed to be just a result of the design, not the opposite way. Design in my opinion is the specification of the needs, maintainability after it is ready, technology used, availability of tools, knowledge base of the CNC maker, the budget, the space and so on. All that must be specified before you make the final drawing. To sketch ideas is OK, but it can not be called a design. Look at it as a conceptual car. Not many of them look like a real, commercially available car and not many of them contain the solutions the designer designed into that conceptual car.

I am not saying it can't be welded but in my opinion welding it is overcomplicating things unless you are very good at it and KNOW what you getting into. Personally I think it is an overkill for the CNC Hartzell described at the beginning. In fact, in my opinion, using a steel frame is also overkill, while it is good to have a rigid, stable frame, aluminum would do just as well in his case. Aluminum is only a problem if the environment, where it is used has a constantly, or often changing temperature otherwise it isn't an issue at all to use softer and lighter material than steel. The total weight in the end will end up at around 40-70kg anyway (depending on the design), so there will enough mass for such small machine. If it was my choice I would definitely go for aluminum frame and I am 100% convinced that it would give him what he needs. Of course, in the end, it is not just about the frame, that's just one part of a CNC, accuracy will be effected by all the other parts as well, including the stepper drivers.

[handlewanker]

Yes, it's hard to envisage someone else's capability and equipment availability.

Personally I gave up working with Mecanno sets when I was a boy.....many years ago.

Having worked in industry for 50 years or so I tend to gravitate to that which I'm most familiar with.

If you're going to work with metal, better to learn how to do it.

Welding metal to metal instead of bolting it together is like gluing wood to wood as opposed to nailing it together.....I expect Hartzel has his own preference.

Being a preferer of welded structures, my suggestions are loaded in that direction.....for what they're worth.......I'm interested to see how the design turns out.
Ian.

[wizard]

Make the bridge flush with the face of the columns not set back to allow full travel across the face of the bridge.

Hope this helps.
Ian.

Ian makes a very good point here. Your current design will critically limit your Y-Axis travel. You might not see that as a problem now but it could be a huge frustration in the future. Ideally your spindles center line should over travel the width of your table. Going a bit farther on at least one side has advantageous for tool changes and setting.

[wizard]

So true, so true...

Not only that, but my experience is that if you think about making a DIY CNC you have to start with your own knowledge base, tools and machines you have and the capabilities around you. It is pointless to sit down and design something if in the end it will need tools and machines which costs a fortune to buy and you can't afford them, or if you lack the knowledge to use them properly. I learned this during my own project, and even though at start I was convinced that I had all the tools and knowledge I needed, I had to buy more and some new tools/machines because what I had was simply not good enough for the task even though I thought from the start that they will be. OK, if you have a fully equipped workshop with a lot of nice machines already, but if not... then you need a lot of good friends who have them, or money to buy and space to place them and to use those tools. This sounds really basic, but I think many people miss this kind of thinking and end up with that castle in the air. There are plenty examples of that on this forum and elsewhere.

This is so true, the concept need to be made a sticky somehow.

I would point out that there is no shame in enlisting the aid of a machine shop or if you have one in your area a makers space. The use of a machine shop might be more expensive than you may think though. Properly machined components though can do wonders for a machine build. As noted though your design needs to be machinable and sometimes that isn't easy to nail down if you don't have shop experience.

[handlewanker]

Hi, I think I did say that..... more or less......if it cost more to make one then buy one.

The whole point of making a machine is to save money........there being many good designs and builds that are complete packages and turnkey with warranties.

Making one to your own design is an exercise in what not to do next time.

If it's a walk down the construction road because that is what you need to satisfy a need to occupy your creative desire then cost is irrelevant.

I think the popular 6040 aluminium built CNC router (moving gantry) on EBAY, going for around a grand and a half, is a good guideline as to how much you won't save by a DIY approach.

You need to be practical and evaluate what it is you need to do.

The only real bit you're going to save on is the frame.....all the other parts are buy in and will cost that much more as individual items with separate shipping costs.

If this is speaking heresy........I did say a practical approach.

The end product, whatever, only has to machine soft materials with a small work envelope, so going to extremes of frame design is not a criteria that needs rocket science and exotic materials to achieve a modicum of success.

I would even go as far as stating that, if the machine was made with wood as has been done very successfully before, the outcome will not be streets behind or lacking in any capability way at all.

Attached is a photo of a very neat moving table CNC router by.......another poster.......and uses wood for the main frame materials......all other parts are buy in off the shelf items.

There is another machine, (photo attached), The Nomad, also in wood, that was a Kickstarter project I think, and is commercially available, but the design is quite interesting as it has the machine completely enclosed in the frame that forms the enclosure.

You have to be crystal clear at the outset as to what you want to have at the end.
Ian.

[A_Camera]

This is so true, the concept need to be made a sticky somehow.

I would point out that there is no shame in enlisting the aid of a machine shop or if you have one in your area a makers space. The use of a machine shop might be more expensive than you may think though. Properly machined components though can do wonders for a machine build. As noted though your design needs to be machinable and sometimes that isn't easy to nail down if you don't have shop experience.

No, there is no shame in using other resources than ones own, but like you say, it may be very expensive, more expensive than expected. Anyway, even if you use a local shop to cut up bigger pieces you can't always run to them every time you need to cut something, so even if you don't need a fully equipped, top of the line workshop at home to make a DIY CNC, I think that most people underestimate the need and the type of tools/machines needed.

[A_Camera]

Hi, I think I did say that..... more or less......if it cost more to make one then buy one.

The whole point of making a machine is to save money........there being many good designs and builds that are complete packages and turnkey with warranties.

Making one to your own design is an exercise in what not to do next time.

If it's a walk down the construction road because that is what you need to satisfy a need to occupy your creative desire then cost is irrelevant.

I think the popular 6040 aluminium built CNC router (moving gantry) on EBAY, going for around a grand and a half, is a good guideline as to how much you won't save by a DIY approach.

You need to be practical and evaluate what it is you need to do.

The only real bit you're going to save on is the frame.....all the other parts are buy in and will cost that much more as individual items with separate shipping costs.

If this is speaking heresy........I did say a practical approach.

The end product, whatever, only has to machine soft materials with a small work envelope, so going to extremes of frame design is not a criteria that needs rocket science and exotic materials to achieve a modicum of success.

I would even go as far as stating that, if the machine was made with wood as has been done very successfully before, the outcome will not be streets behind or lacking in any capability way at all.

Attached is a photo of a very neat moving table CNC router by.......another poster.......and uses wood for the main frame materials......all other parts are buy in off the shelf items.

There is another machine, (photo attached), The Nomad, also in wood, that was a Kickstarter project I think, and is commercially available, but the design is quite interesting as it has the machine completely enclosed in the frame that forms the enclosure.

You have to be crystal clear at the outset as to what you want to have at the end.
Ian.

The aim of DIY (anything) is not to make something cheap. At least in my opinion, the theory of "if it cost more to make one then buy one" is wrong and can only be applied to pure commercial activity.

Although in my previous post I was going to mention the wood or MDF as a possibility, and it may even work for some small jobs, I think I'd advise AGAINST. Even if it may initially work well, I think that over time, and probably very fast, the problems will show up. Wood, and everything built in wood, changes shape too often and is not predictable. Wood expands and softens when it is moist and then, when the air is dry, it shrinks. Even in a well tempered home, the change of seasons will cause problems. Just look at it as a wooden floor. Ever tried to build a house or a room with the floor going exactly from one wall to the other? Because if you lay the floor during the dry winter season and then the summer comes you are up for a nasty surprise... That nicely laid floor will no doubt look like the high sea, or will push the walls apart...

Aluminum is also expanding and shrinking with changing heat, but moisture has no effects on it, and the affects of heat does not matter in a normal home or workshop, where the temperature is fairly constant, so while like I said, aluminum is an excellent material to make a DIY CNC from, I would not build one out of wood, fiber board or MDF, at least not if accuracy and precision matters.

Never the less, covering the CNC, at least partially, is a good idea because it allows a tidier environment and eliminates the need of personal protection when the machine is in use. The second picture shows a completely covered CNC, which isn't a very good idea unless there is also a very good cooling system, not only for the spindle, but also for the stepper motor. Remember that steppers can get very hot and if the machine is completely covered then that heat will increase and that isn't very good. Commercially build CNCs are many times covered but they are also properly ventilated to prevent overheat. I have also designed my CNC with protection cover and from the start I planned to cover it completely but after some thinking I never implemented the top cover to allow the heat to rise. Since all the four sides are covered high enough, and the bottom has also a cover, I can use it in any environment, at least as far as flying debris is concerned. Very little is leaving the boxed area, so currently I am using it in my home office. Of course, cooling fluid can not be used there, but in my case, I will not be using any so that's not a problem. I am also not doing any wood work, so fine dust is not an issue.