Platen Design

[datbuko]

Hello,
I am a recent graduate working at my first job. I have been assigned to redesign a flour tortilla press. The problem that I am having mainly deals with heat expansion of my frame/platens. We use a ridged 4 post frame to keep the upper frame stable. However, due to thermal expansion, the post are expanding and transferring a side load to our hydraulic cylinder, thus causing a lower life expectancy. Another major issue is that since we are making flour tortillas, platen distance is very critical.Due to the 42x50" platens being locked down via bolts,I believe this is causing local deformities within the plate, thus translating to deformed tortillas. Any help would be very appreciated. How do they overcome this issue with molds?

Thanks,
David

 

[EngineerTex]

A picture or diagram of your current setup would be helpful. I'm having a hard time understanding what you're describing.

Engineering is not the science behind building. It is the science behind not building.

 

[MikeHalloran]

Molding machines use very thick, very rigid platens, and very long guide bushings. I'd bet your machine is nowhere near as rigid. E.g. a molding machine with a 42" x 50" platen would be the size of a locomotive.

Absent that, perhaps you could put spacers around the edges of your platens, so that the platen faces are held apart by the thickness of a tortilla at least.



Mike Halloran
Pembroke Pines, FL, USA

 

[datbuko]

This is a picture of the my current idea.

 

[datbuko]

Here are the details for the bolts. I am attempting to make this a "floating plate" design, so that when the plates expand due to heat (400 deg F) they do not warp due to thermal expansion and bolts.

 

[datbuko]

Here is a picture of my floating guide post design.
Any input/suggestions are very much appreciated.

 

[MikeHalloran]

Your guide bushings are way way way too short to do any good.



Mike Halloran
Pembroke Pines, FL, USA

 

[datbuko]

Mike,
Do you know where I can source longer bushings. Most places that I have seen online go up to 5" or so.

 

[Pud]

I assume the platens are heated? If so, are they insulated from the frame? Injection molding tools which have to run hot (e.g.+100 deg C) have insulating plates on to keep the platens cool, and to reduce heat loss from the tool.

http://www.tynevalleyplastics.co.uk

 

[dhengr]

Datbuko:
My goodness, you guys sure can make complicated drawings look well thought out with all this fancy 3D CAD these days; without nearly enough thought about what you are really doing or trying to make. There seem to me to be many more questions raised in your in your post and drawings than just thermal expansion and bushings. Longer bushings will just distribute the load and wear over a larger length, not eliminate the loading or solve the problem with your arrangement. Your press looks very complicated for what it is intended to do. And, complicated does not always mean good, clean, long lasting design. You should strive for clean, simple, easy to maintain and adjust design, not complexity. Why do you actually have lat. loading on the guide posts? You can’t guide something which expands off of a structure that doesn’t expand at the same rate, at least not easily.

Just out of curiosity, what is the compressive strength of flour tortilla batter prior to and during cooking, less than 8000psi? That ram and structure looks like it would make diamonds out of coal with a bit higher temp. Other than wanting the platens to remain fairly flat, and align laterally, what are your ram force req’rmts.? What tolerances do you really need on vert. and lat. alignment, and platen flatness? I’ve never used a micrometer on a package of tortillas before. What are the cook (hold top platen down), unload, reload, cycle times? Do you have to get under the top platen to load and unload the bot. platen?

I would guide most of the motion with some fairly simple structural guides, maybe cam rollers running on the upper structure, to engage four tapered holes (top platen) and pins (bot. platen), one at each corner of the platens. I would provide some hard stops vertically on one platen for thickness and then pressure relief on the cylinder when this contact is reached. I assume both of the platens are heated so the thermal expansion problem is really not btwn. the two platens. I would like the outer edges of the platens to touch first, and then the center, as the vert. stops come into play. The top platen could be a fairly simple, flatter, fairly rigid, grid work of stiffeners. This problem is basically a structural elastic and thermal matching problem.

 

[datbuko]

dhengr,
Here is one of the many fea's that I did for our upper structure. A standard 7" flour tortilla while press gives roughly 103 psi and we are pressing 25 tortillas per cycle. As for tolerances, the lateral is not a large issue unless the plates when compressing slide, thus giving an oval tortilla. However, the platen flatness and vertical alignment makes a very big difference in the tortilla size. A 0.005 difference .25" difference in the diameter of the tortilla.
To give a little background on how it works, we take a round raw tortilla dough and press it at around 300-450 deg (depending on customers dough formula). The heat is used in order to keep tortilla from retracting as well as allowing it to release from the Teflon belt. Then it is cooked in a separate oven.
Our press is unique in the fact that it is a continuous motion press head, meaning the press travels back an forth on rails with the speed of the belt in order to continuously feed tortillas threw. We run approximately 21 cpm with a .5 sec dwell.
For overall thickness we use the stroke of the cylinder as our hard stop.

Pud,
We were using 1" thick glastherm, however to keep weight down I decided to go with Pyropel.

 

[ABritTurner]

When first reading this I thought it was a wind up. I still think it might be a wind up, but if it is serious I will add my two cents. dhengr raised some interesting issues and I tend to go along with his way of thinking.

I once manufactured a rolling machine where the tolerances specified by the designer were 0.0004" that translates to 4/10th of a thou. The designer made a mistake in the tolerance on his drawing ans wrote 0.004" which give a latitude of 4 thou not 4/10 of a thou. These tolerances were on the guide post of this machine which were very much like your machine.

He had designed these machines before and there was always a problem, the wire when rolled tended to spiral and that caused him problems so he invented an anti spiral device which worked after much fine tuning.

He realised his mistake just as I finished the manufacture of his rolling machine and then proceeded to fit his anti spiral device. A year later he was still struggling with the anti rolling device. Eventually he threw the anti spiral device away and the net result was that he rolled perfectly formed wire without any spiralling. In short his tolerances had always been to tight, we actually manufactured the machine with around two thou float for the guides and these were the optimum.

Moral of this tale.
Designers tend to think tighter tolerances are better, a common mistake also made by the layman. if the machine you have is distorting it may be because you don't have enough lateral movement, try larger allowances on your guides, give them double what you have now and see if it works. It cannot be any worse than what is happening now.

The same thinking applies to engineers who decide to alter datums, if a datum has changed something has changed and it could simply mean a worn thrust bearing. The same thinking is with hardened parts, people tend to go with the hardest material when a slightly less hard pin would do better. Or necking of stress loaded bolts, the necking actually allows the bolt to bend and thereby reduces stress and fracture.

 

[okeng]

Design Rule #1:

Slideth not thy parts; rather let them rotateth, that it might go well with you and your children.

Seriously, unless it's really important for the two surfaces to move linearly, making things that pivot is generally more robust, simple, and reliable than making things that slide. Put the upper platen on a long arm, or 4 short parallel arms if space is tight, and then make sure that those pivot points are the ONLY things that require precision. Make everything else loose and rattly.

I do understand, of course, that if the customer has an expectation of what it will look like, then you're stuck optimizing the slides.

And remember--sometimes it's better for inadvertent flaws to just be sold as features. Your customers can charge extra for oval tortillas, because they're more....elliptical. Sure.

 

[MintJulep]

causing local deformities within the plate

Or, they could be thermally induced.

Your stiffening ribs are also cooling fins. No way this thing will ever be flat. Ever.

This is a fundamentally flawed design and it will never work the way you want it to. After much tweaking it might work okish.

Use a roller.