Automatically actuated grill block - Mk3
 

So, I've tried this before, and each time its become a pain in the butt in some way. This isn't abnormal in the product development process, so we keep refining the design until we come up with something better that works and meets all the design criteria. For kicks, lets go over the past designs.

The first automatically actuated grill block was extremely simple in design, but not structurally strong enough. It was made mostly of coroplast and when installed on the car it flexed too much. All of the flexing essentially eliminated the ~1" throw of the actuator I was using.

http://ecomodder.com/forum/attachmen...1&d=1232564425

http://ecomodder.com/forum/attachmen...1&d=1232564425




The second design was a lot more rigid, and designed with scalability in mind. It had an aluminum frame, with coroplast blades that had steel rods in them for reinforcement and pivoting on. However, it ended up being a pain to make and put together. It was quite time intensive building it, and it still had its issues with the pushrod setup binding.

http://ecomodder.com/forum/attachmen...1&d=1354131785

http://ecomodder.com/forum/attachmen...1&d=1354131785



This brings us to the Mk3 design. This version is also designed with scalability in mind, and the frame is also quite rigid. However, I didn't want to spend time building it, so much so that I don't even want to use tools to assemble it. So, this grill block is going to completely snap together. Sounds lazy eh? Sounds brilliant to me. :) So, I'm posting what I have thus far. Darin has been bugging me for a few days now to start a thread about it since I started telling him about it. It is not complete, but it is a fair ways along. The design is nearly 100% 3d printed. The actuator will bolt to the assembly of 3d printed stuff and you'll probably need a pushrod or two to finish things off. That is the hope at least.

This design is primarily made for my 2000 Honda Insight. It has a very short lower grill. The opening is only 3" tall, so I only needed a single blade. The grill block when open is currently shown as 2" x 6". I'll probably end up making it a bit longer, but its about as tall as it is going to get.


Here is the blade. You can see its made of 3 pieces. If I need to make it longer, I can make the pieces longer, or add additional pieces that snap together between the end pieces. I've printed these pieces and tested out the snap fit and its surprisingly strong.

http://ecomodder.com/forum/attachmen...1&d=1478621034



Here is the frame. I have not been able to print and test these snap fittings yet. They're very similar in design to the blades though and I'm pretty sure they'll be fairly strong. If not, they can be tweaked.

http://ecomodder.com/forum/attachmen...1&d=1478621216



Here is the assembly as far as I have it so far. The blades fit into the slots on the frame, and there is another clip that holds the blade in. It is 100% snap together.

http://ecomodder.com/forum/attachmen...1&d=1478621216

http://ecomodder.com/forum/attachmen...1&d=1478621216



The next step is designing the mount for the actuator (a power door lock actuator), and developing the linkages between it and the blade. I already have some ideas, I just need to model them up. I'll be updating this thread as I get more designed.

As always, I'm open to ideas and feedback that you guys have.

Dang! That's pretty neat! Keep up progress, I need one soon. :)

Take my money.

Hah, I figured the Insight guys would like it. If it works out well, I hope to redesign it for the Prius as well. It would be even more useful on that vehicle as it warms up slower than the Insight.

I'd like one on the Leaf, I'd say that I also want on in the TDI, but at 1.5 tanks / year I can no longer justify modifying it.

Subscribed:

Nice work, make one for my GEO so I can do away with the chaps & spikes look... !

Where do I sign up for one? lol

http://ecomodder.com/forum/showthrea...ers-31585.html

Want to buy my prototype? Been hanging in the garage since graduation. Worked on the bench but never put it in a vehicle.

Doax, you might want to "design-in" a steel rod *inside* the two pivots rods, or alternatively slide a metal bushing over them to both reduce rotating friction and provide physical strength.

Interesting idea Old Tele man. I do want to keep the design as close to 100% 3d printable as possible. I don't think friction will be an issue. The blades are very light weight, and all the joints are designed with a bit of play in them. Those power door lock actuators also have quite a bit of oomph. I do like the idea about increasing the pin diameter to strengthen it. I had considered it, but I'd like to try it out as is first. If I increase the pin diameter it'll make the blades slightly unbalanced due to how they have to be printed on the machine.

Another great thing about this multi-piece design is if I do need to increase the pin diameter or make any other changes, I just reprint those end pieces, unsnap old pieces, and snap the new pieces together and bingo done. No rebuilding the entire grill block.

I do NOT know exactly what the torsional loads might be at highway speeds (+50% for headwinds!) but end activation could "un-snap" your connections (unless your plan is to ultimately glue them together). A center activation would greatly reduce that torsional load, whatever it is.

Again, thought provoking ideas Old Tele Man. :thumbup: Perhaps I should bias the blade pivot point so that the wind holds it shut at speed? With the blade open, it won't take much force to keep it in that position. The door lock actuator has some friction to overcome with the gearing inside of it, so it will help hold the block in place at what ever position its at, but a little help never hurt.

Here are a couple more pictures. I got the actuator mounting bracket designed.

http://ecomodder.com/forum/attachmen...1&d=1479305458



Here it is with the actuator shown. This more accurately shows the scale of things. The grill block is not that large. The entire assembly is only 3" tall and 9" wide currently. It will get larger to add some sort of mounting provision. I just haven't gotten that far.

http://ecomodder.com/forum/attachmen...1&d=1479305458



Next step is designing the linkage between the actuator and the blade. This has been proving tricky as I'm running out of room.

At this point, wouldn't a rotary actuator actually be better/easier to interface?

Red Dot-Red Dot Electric Operated Rotary Sealed Servo Actuator 12V 4-Pin - 71R3576 - RD-5-9063-0P-71R3576

It is quite possible that a servo would work better for the application, especially if it were smaller in size. The reason I keep going to the door lock actuator is price. These door lock actuators are $5 shipped. How cheap can you get servos for? If you can link me to a good priced one I'd definitely be interested. The whole thing will be arduino controlled, so controlling it is not an issue.

20KG Metal Gear High Torque Wide Angle Digital Waterproof Servo for RC Robot PZ | eBay

Claims 20kg force on a 1cm arm. Derate by 50% and double the arm length and you're still looking at 5kg, that should be enough, right?

Waterproof, all metal geartrain, works on 5v. $14.89 + $8 shipping :(

Also:
Towerpro MG958 55g 20KG High Torque Digital Metal Gear Servo | eBay

20kg, metal gear, not waterproof, $14.83, free shipping

Lots of pretty cheap options on ebay for servos nowadays.

I actually have some radio control servos kicking around the house. I bet even those little ones have enough power to actuate the grill. I'm not sure about holding it in place... but we'll see. I can design it so the wind speed holds it shut. I like the servo idea though. Also, I looked on amazon and you can get 2X micro servos for about $8. Its WAY smaller than the door lock actuator and would be much easier to integrate into the design. Thanks for the idea!

https://www.amazon.com/TowerPro-SG90...ds=micro+servo

"It takes a COMMUNITY to design a grill block."

It certainly helps to have a wealth of knowledge kick in their ideas! Thanks guys. :thumbup:

I dug through my radio control stuff and found a micro servo. Its really quite small (mechanical pencil in the picture for scale) but does have a pretty great amount of holding power. I was a bit surprised about that. The gearing must be a pretty tall ratio. Anyways, I think I should be able to mount it directly inline with the lever arm that actuates the blade.

http://ecomodder.com/forum/attachmen...1&d=1479391255

Daox already answered this for me but asked I post my question here:

Quote (Originally by RoadCyclist)---
Hey, I was reading you Arduino grill block thread and was wondering: How are you programming the Arduino? Is it linked to the engine computer? Also, how are you powering the servo and MC?

I'm asking because I have an Arduino and can do some simple tricks with it, and was thinking about copying your design if that's all right with you, assuming it works out.

I just started looking the the arduino servo library the other day. However, it seems very easy to use. First, you setup a servo in the code. Then, at any point you can tell the servo to go to X degrees. Its really that simple. Here is an example code that moves a servo based off input from a potentiometer. My code will be very similar, except that it'll fully open or close the block based off of the temperature of the coolant. At this point I don't plan on partially opening the grill block.

https://www.arduino.cc/en/Tutorial/Knob

Once I get done with the mechanical stuff, I will be moving into the electrical and I will post my code when I get there.

The arduino will be tapping into the coolant temperature sensor signal so it knows when to open and close, but that is about it other than hooking up power and ground wires.

As for powering the servo, it should be able to be run off the 5V power regulator onboard the arduino. The specs on the regulator say it can handle up to 800mA, and the servo specs say it uses 200mA. So, we have plenty of head room for the arduino to do its job.

A. But how are you powering the Arduino itself? It needs power too!

B. It wouldn't be too hard to divide the servo activation into steps, i.e. 1/2, 3/4, 1/4 open, depending on the input value. Again, I'm no expert here, but that's my opinion based on what little experience I've got.

Thanks for sharing your research, by the way! I really am looking forward to your progress.

The arduino will be powered with 12V switched power from the car. Most versions of the arduino have a power regulator circuit built into them, so it can easily accept 12V input. No filters (capacitors) needed as they're built into the regulator circuit.

Quite true, it would not be too hard to program the servo to open incrementally. However, I don't see much advantage? As I see it, ideal operation is to wait until just before the electric fan turns on, then open the grill block. This ensures the radiator is completely full of the hottest coolant possible. That means our bang for the buck of airflow will be the greatest. I'll drop the coolant temp down to just above where the thermostat closes, and then close the grill again. It will cycle open and closed. I don't think it'll be too often, but testing will tell. If it is actuating too often, adding variable grill block logic would probably be advantageous. If anyone sees flaws in my logic, do tell! :thumbup:

Good work! It looks like it is drawn in Google SketchUp. Is it?

Thanks seifrob!

The software I use to draw up all my stuff is SolidWorks. I use it professionally daily so its a lot quicker for me to do it in that than Sketchup. I did try Sketchup once. It seemed pretty nice for a free 3d software package. But, since I already know Solidworks, that is what I stick with.

If you find the micro servo dose not pack enough punch,then consider using a spring for close and setup the servo as a winch. Insted of a larger servo

Winch setup

Opt1 disable the potentiometer (as per instructions below)
Opt2 Relocate the potentiometer to the grill flap (this option would delete the need for limit switches)
-crawler-tech-how-build-winch.html
Gumby Stay Flexible

Interesting idea, thanks Gumby. :thumbup:

A few gotchas for you to watch out for:

The arduino reg generates a lot of heat and may not like extended periods at 14V. Might be better to have a pre-regulator to knock it down to 8-9V and kill more noise too.

By using proportional control of the servo you will prevent cycling and excessive open/close cycles. I would choose an opening amount based on the rate of rise of the coolant temperature and modify this in a closed loop to get equilibrium rather than flopping around like a dying fish between extremes.

You might also consider the cooling fan as an additional input to adjust the strategy - if the fan is running for longer than x minutes then open up more and conversely if it is off for y minutes then close it up a bit.

Ambient temperature should also be used to modulate the opening rate / amount - I would not slam it open if the air temp is -20.

You could also use an analog input connected to a variable resistor to change the reactivity (gain) of the loop for tuning.

Simon

Good point about the 14V. However, the vast majority of the time the servo isn't going to be in use. So, the load will primarily just be the arduino which doesn't really pull much power (~45 mA per a quick google search). However, I will keep an eye on the regulator temp. I'd like to keep this as simple as I can. If it becomes an issue, I'll look into changing it.

I like the idea of closed loop control to create equilibrium. Perhaps that'll be stage 2 once I get it installed and working.

Ideally, I'd also like to come up with a way to close the grill block when the car is shut off. This would help retain some heat in the engine for a bit longer. If anyone has any ideas on that, I'm all ears. I'm not sure how much this would really be super useful though if it takes much more complication to implement.

I've been printing as I've progressed to test out the pieces, but I figured I'd show some of the images with you guys. Its not vaporware, I swear! :)


Here is the initial blade design. The blade is 1/8" thick here. I later beefed up the pivot point to 3/16" diameter. I think 1/8" would have been fine, but 3/16" will definitely be plenty strong.

http://ecomodder.com/forum/attachmen...1&d=1479486968




This was the test piece for the frame. It is surprisingly sturdy once snapped together. I also tested the blade clip which needed a little tweaking. The snap was a bit too tight. The other clip you see on the far right was for the power door lock actuator mounting piece which has now been eliminated.

http://ecomodder.com/forum/attachmen...1&d=1479486968



These parts were just printed last night. The blade clip is a new design and it works beautifully.

http://ecomodder.com/forum/attachmen...1&d=1479486968



The servo mount also works very nicely as well. There isn't a ton of extra room though! :) So far I'm really happy with how the design is coming together.

http://ecomodder.com/forum/attachmen...1&d=1479486968

I love it when a plan comes together... That looks great!

You will have to program the idle time off (no power) . The servo uses power to hold position.
--- for shutdown
Add a capacitor the power the unit after shutdown. program when the unit sees the engine shutdown to close the flap

Thanks RoadCyclist.

I got the rest of the bits all printed up last night. They all go together pretty well. The blade doesn't fit up perfectly. I'll get that sorted out though.

http://ecomodder.com/forum/attachmen...1&d=1479573823


Next up I'll get the servo hooked up.

Does the servo use power to hold position, or only when it gets pushed out of position? I'll have to hook up an ammeter and see how its been programmed.

Most servos have plain spur gears but the ratio is quite high, resulting in a significant torque required to move them without power. There is no power used by the servo motor when in position, but the electronics uses a small amount. The pulse position tells the servo the desired position, which is compared to the internal position and the potor actuated till they match. If it is moved out of position it will resist and try to move back.

For closing after ignition off you will need a really big capacitor to provide enough power to close the grille, probably in excess of 100,000uF. You can use a resistive divider and an analog input to monitor the input voltage before the regulator to detect power off.

Simon

The load will be speed based bell curve accentuated by the off center pen .
Anxious to see the results of your testing. It will answer some of the questions I've had in my idea .http://ecomodder.com/forum/showthrea...all-33611.html

I got it fully assembled and tested out, and shot a quick video. Its running a slightly modified version of the 'sweep' program that is an example of how to use the servo library. Obviously, this program just sweeps the servo back and forth.

I did hook an ammeter up to the servo while it was in use. Sitting idle it used 10mA. While it was moving it used 60-70mA. However, the program it was running does not actuate the servo at full speed (which would be fine for this application). I imagine full speed would draw a lot more power.

https://www.youtube.com/watch?v=uQS91D-95Po

Have you tried "cycling" it under a wind load, such as a hair-drier or desk fan?

Nope, I have not tried that. I'll fire it back up today and see what I can do. I don't think I have a blow drier though... I know I have a fan but that is going to be a fairly low wind speed. Good idea though. Perhaps I'll just try to hold it with my hand and see how much oomph it has. I'm pretty confident that it'll be okay. There is a bit of travel that I didn't use in the servo. It has 180 degrees of rotation and I think I'm using about 110 degrees of that. If worst comes to worst, I can use more of that throw that for a bit more mechanical advantage. Also, as you can see in the video, I stop the servo actuator arm parallel to the push rod. This gives me tons of mechanical advantage since the push rod is pushing straight into the servo pivot point vs trying to rotate the arm.

It doesn't look from the video like it is closing all the way. If true, you could lengthen the wire arm some.

Speaking of the wire arm, I'm not sure it would stay straight trying to hold a flap closed in high speed conditions.

(My 2 cents)