Peer review - Mechanical 3D models

Good morning,

In the last few days I have completed the design of the mechanics relating to the design intended for the first sheet metal prototype.

In the CAD folder of the project repository, you can find all the 3D models in neutral format (STEP) of the mechanical part.
In the PARTS subfolder, you can find the individual parts to be made, while in the ASSY subfold there are the general assembly “A00” and all the related subassemblies, some of which without the small parts to make it easier to view.

Let me know what you think about the design of the models. Obviously, any form of contribution, from identifying critical issues to suggestions on how to improve the current design, is extremely welcome.

The criticality of the thrust bearing system is currently well known and will be reviewed shortly, but as already mentioned, any suggestions also on this topic are more than welcome.
Thanks everyone for the support!



Dear Patrizio,
I forwarded your request to my colleagues (mechanical engineers) to gather their feedback.
Hope to support you soon.
Feel free to contact me if you have specific needs

Dear all,
I see the 3d model and the sheetmetal assembly this evening for the first time.
For my experience we can simplify the mounting operation and work to reduce fixing screws and parts.
For example, you are using traditional bolt and screw concept, but with sheetmetal for me it´s smarter to use cage nut (if necessery one direction adjustment) and blind rivet nut (remains fix with sheetmetal and make easier the assembly process and manufacturing).
Tomorrow if you want, i would have pleasure to show you some proposal…bye

p.s.: you use micro switch for safety reasons or start and zero function? And if i´am not mistaken, there are no mechanical limit switches or dumper at the end and start of the stroke?

Just updated the main assembly ( A00_OB_200401_r03) with a new thrust bearing system, and a few fixings suggested by some great guys from the Cern’s mechanical engineering group. I still have to check every connection and replace some washer that is causing some interference at the moment.

@Vito I will be more than happy to have their feedback!

@dezi.marco it will be nice to implement some faster or simpler solution to speed up the production and assembly stage.
The micro switch are used for both start and zero function, to check the presence of the ambu bag, to check if the clamp are closed (some of the locking mechanism at these stage are just related to the first prototype, they will be replaced with a toolless system). No there isn’t any mechanical limit to the stroke. Did you think it should be implemented?

Dear Patrizio,
I forwarded your request to my colleagues mechanical engineers for their feedback.

At disposal

Best regards

Dear Patrizio,
my colleagues raised up some main concerns about the current design of your prototype. Maybe you are already aware of some of them, so you can jump over them. Following a synthetic list:

  • the friction between the V-mechanism and the balloon is supposed to consume and damage the balloon; a possible solution is a rolling free cylinder embedded in each of the sides of the V-mechanism (maybe, not easy to design and manufacture);
  • due to V-like the mechanism, the balloon can shift upward; thus, its deformation might not be fully controllable;
  • the thickness of some parts made in metal sheet should be increased for safety issues; they are supposed to properly work over many repetitions;
  • if possible, use the same size for all bolts and nuts, to make purchasing and stocking easier.
    Additional details concern autonomy and air flow-related sensors (mostly electronics):
  • will the ventilator be provided with a back up battery?
  • how will the air-flow (e.g., flow rate, pressure, O2 percentage) be controlled?
    I guess these issues are relevant (or mandatory) for this kind of biomedical device.
    Furthermore, I assume that you designed a redundant V-like mechanism and related sensors for safety issues, thus simpler (i.e., one-actuator) mechanisms are out of the topic. Anyway, I would suggest you to report also specifications you defined and met to design this first prototype. These information will be helpful for the reviewers to better understand the rationale underlying your choices.
    Hope these comments are helpful. Feel free to contact me if further clarifications are required.

I agree with Patrizio about design specification, and if possible, would be usefull the datasheets for the items like encoder, support unit, bush or bearing…if we want do a real review we have to know how you want control ad use the unit, and if you have some calculation or specification (motor torque for example…and the relative axial force on the screw unit).
And now also some question:
why you use 3 encoder? the first on the motor, and two different type on axis V mechanism. Redundancy?
The microswitch on the squeezing plate for me doesn´t work correct…the balloon deformation on the micro thin lever maybe create some sticking or damage and bend it…
For the mechanical limit to the stroke, about my previous post, depends how you control the motion…if your PI controller system overshoot during operation (for example for a noise on encoder signals or some extra torque due to some friction…) you can damage the microswitch, pushing against it with full load. If you have mechanical damper system, adjusted according to microswitch position, are safer…
For me you have too much seeger on one axis…you can eliminate the red ones.


Or if you want modify for less components for me you can use a configuration with the two green and add a key on left side for the rotation of axis with the squeezing plate…

The idea seems very valid to me, would you have any suggestions on how to make it with possibly standard components?

To be honest this is the biggest doubt I have since the beginning of the project, we will find out very soon since the first prototipe in nearly ready.

This I believe is indispensable, it will be very nice to have some contribution on this point.

Well noted for the next step, I guess is a great idea.

Yes, it will.

Yes, they will. At the present the biggest concern is to find an O2 affordable sensor, the back up plan is to use the tabular data of the ambu bag.

Thnaks for your feedback!

@Patrizio [IMPORTANTE]: sono in linea con un potenziale partner industriale; puoi contattare Solinas perchè ci contatti quanto prima? grazie

I’ve uploaded the encoder’s datasheet, and calculation sheet feel free to check them and and possibly corrected them.

Yes they are supposed be for redundancy, the two different model is just for the testing stage, then a single model will be used. There is also a custom encoder in development.

I’m aware of this, we will protect and prevent the sticking between micro-switch lever and the ambu bag with an plastic sheet. If you have a better solution for the production stage, please feel free to propose and design it.

I’ve put the “extra” seeger in order to prevent any misalignment of the bended sheet metal, and obtaining a better hinge system, I’ve some concern about the precision of sheet metals parts…

Thanks for all the ideas and tips!!

it is the first time that I participate in a forum, so this for me is also a test. I could help you in mechanics and numerical simulation of your apparatus. I downloaded from repository the CAD model importing it into Creo Parametric 5 (PTC). There are some import errors and I am evaluating how serious they are. I can make you some proposals on numerical simulation?


Hi tojfl, for sure would be extremely useful to evaluate the actual design by a FEM analysis. Let me know if you would like to setup a meeting to check the system constrain and loads.

Please consider last update (A01_OB_200401_WIP_NEMA_17_23.step) as the reference for your analysis.

It would be a great contribution to the project!

I made some proposal on sub-system assembly.
a05_ob_200403_dez_asm.7z (1.7 MB)

One plate for motor mounting with blind rivet nut M4…the previuos 4 bar link it is not so torque resitant…

And I made 2 plate with threaded holes to fix the screw support and nut…

P.S.: check your assembly , the contact face of nuts in some case maybe on wrong way…and for extra seeger, the more you’ll put more tolerance problems you will have


Yes, a meeting i very useful. In what way? Skype? From my side (SSSA) it is possible also to use a “room” by WebEx, but I need to book.
Unfortunately the import errors are present also for A01_OB_200401_WIP_NEMA_17_23.step, and as yesterday for whole assembly I was unable to solve them. By exporting, do you have any options? By reading the source step file “Autodesk Translation Framework v8.12.0.” is reported as origin: do you use AutoCAD? In any case, the import process only partially fails and my imported assembly is on with some wrong imported feature. So, if you don’t have time don’t worry about it.


@ dezi.marco i’m looking at your proposal, the motor flange seams very good to me.
@tojfl at the moment I’m using Fusion360, in the next days I should switch to Catia, wich format Creo could reads the easiest?

Probably with Catia will go better, but it will be necessary to try. For the moment, if you have time, you could export to IGES.


Pushed the .iges, please try if it solves the I/O problems…

There seems to be only one import error, on D1.ASM. The situation has clearly improved.


Hello everyone, sorry if I have not updated the models with the proposals that have emerged so far, but I’m reviewing the thrust bearing system, since, the manufacturer is unable to supply the system because the company seems to be blocked. I’m thinking about using a double angular contact bearing, and turning a custom flange to mount it as it is currently in the CAD model.

Another idea is that since I have to turn a flange, I could center the bearing on the face of the NEMA, and replace the elastic joint with a rigid one.

What do you think about it? Do you have any suggestions on this topic?