Specialty Chai Latte Machine

Specialty Chai Latte Machine

Sponsored By: Prince Gupta | Anand Parikh

In Collaboration with Rady School of Management

Figure 1: First prototype's final design of the Specialty Chai Latte Machine

Objective:

This senior design capstone project is the beginning of a start-up that is seeking to launch a product onto the consumer market. Recognizing the incredible popularity of automated coffee and espresso machines, project sponsors Prince Gupta and Anand Parikh are taking on the challenge of creating the Speciality Chai Latte Machine: a would-be new contender in the automated beverage-making arena.The inspiration behind this device draws upon the extreme convenience offered by machines crafted by companies such as Keurig or Nespresso along with a desire to simplify the normally arduous process of making Chai Tea.

The objective for this project can be split into three parts: 

1. Construct a fully functioning prototype of the Specialty Chai Latte Machine 

2. Obtain customer feedback regarding the design of the prototype and the taste of the tea

3. Begin the brainstorming process for a second, more refined prototype 

The Challenges in Making Tea:

In India, Chai Tea is the equivalent of coffee to Americans. Everyone drinks it as a breakfast item. However, the process of making Chai is considerably involved and requires constant supervision. In particular, the risk of frothing and the overall cooking time are the two most significant headaches associated with the Chai process.

Frothing:

Frothing is another term for "boiling over": where the liquid contents of a pot seemingly bursts out of its confines. This phenomenon is directly caused by evaporation. More specifically, as a liquid mixture heats up and releases water vapor out of a pot, lipid and protein residue is left behind. As more water is evaporated, an increasingly thicker lipid-protein film becomes deposited until water vapor starts being trapped underneath. Naturally, the water vapor will continuously seek to escape and thereby lift the entire lipid-protein layer upwards and out of the pot. 

Cooking Time:

On average, cooking Chai in the traditional manner takes approximately fifteen minutes.

Prototype Overview:

Two views of the prototype assembly are shown above: the left being an isometric and the right, a side perspective. Aluminum extrusions were used to create a cubic frame within which the various components of the prototype will be inserted. This choice for construction takes full advantage of high modular capacity that the extrusions' T-slots allow. In this manner, various internal configurations can be tested upon prototype assembly and can eventually drive the final design of the Specialty Chai Latte Machine's redesign. 

The overall dimensions of this prototype frame is approximate 22 cubic inches and features four main mechanical components:

1. Solid dispensing mechanism

2. Liquid dispensing mechanism

3. Combined stirrer and carafe lid apparatus

4. Tubing and pump system

Solid Dispensing Mechanism:

Functional Requirements:

1. Dispensers must be easily removable for cleaning

2. Easily Cleanable

3. Accurately measure and provide the proper amount of ingredients for 1,2,3 or 4 servings (tea = 1 tsp/serving and masala = ¼ tsp/serving)

4. Containers must store extra ingredients so that the user may make many servings before having to refill the dispensers

5. Design must be easily adjustable in the event that tea / masala ratios must be changed per serving

Liquid Dispensing Mechanism:

Functional Requirements:

1. Easily detaches from and attaches to the machine’s main body

2. Easily cleanable 

3. Dispenses proper amount of water and milk

Combined Stirrer and Carafe Lid Apparatus:

Functional Requirements:

1. System must be able ensure that the ingredients are mixed into the brew properly rather than being stuck to the sides

2. Apparatus must easily mate with the rest of the machine

Tubing and Pump System:

Functional Requirements:

1. Transfers milk and water from the liquid dispensing contains into the carafe

2. Easy to clean

Summary of Performance Results:

Prototype performance as it applies to the SCLM is most aptly explained by discussing the tests performed on each individual component since the prototype is built by consolidating separate mechanisms. Furthermore, since much of this project is focused on mechanical design, designers’ expectations will be discussed in lieu of “theoretical predictions” as a baseline against which actual prototype components can be compared against.

Electronics:

Designer Expectations:

The electronics circuit is designed in a manner such that it can control all aspects of the machine through an Arduino microprocessor. This circuit must be able to receive communication from temperature and flow sensors as well as give appropriate commands to the motors and pumps such that the solids and liquid dispensers operate in proper sequential order. As a result, the Arduino-based circuit must be able to:

·         Power on induction cooker

·         Adjust induction cooktop’s strength

·         Turn on and off all DC motors and control them using PWM

·         Turn on pumps

·         Monitor flow using a flow meter sensor

·         Open and close valves

·         Read values from a temperature sensor

·         Initiate tea-making process at the touch of a button

·         Initiate cleaning cycle at the touch of a button

Test Conditions:

A completed prototype with electronics fully integrated was used to test the electronics aspect of the SCLM. The following figures (4.1 to 4.3) detail the electrical setup of the finalized prototype. For protection against any potential splashes or liquid and general contamination, the microprocessor and all protoboards are secured in a Tupperware container mounted in the back corner of the SCLM. Should any modifications need to be made to the circuit, the Tupperware lid is easily removed. A panel of MOSFETs is also mounted at the backside of the SCLM and controls the delivery of power to the SCLM’s various DC motors and pumps. Electronics testing consisted of careful monitoring of the Arduino Uno’s Serial Port Monitor throughout a complete tea-making cycle.

Results:

Electronics performances for the final prototype are outlined below according the specific components:

1.      Induction Cooktop: Relays were attached to the induction cooker switches and linked back to the Arduino. Using the relays, the Arduino was able to turn the cooker on, off, and adjust the power without any issues.

2.      Flow Meter: The flow meter came with a manufacturer’s code from the supplier for reading and processing with an Arduino. Tests have demonstrated accurate readings of the flow rate and more importantly allows precise volumes of liquids to be dispensed.

3.      DC Motors: The DC motors were all tested individually with an external power supply to ensure proper function. During the tea-making trial, all motors functioned as expected. The PWM commands were visually confirmed to be operational by examining corresponding increases or decreases in motor speeds.

4.      Pumps and Valves: No issues were detected during the brewing process. These two components were perhaps the most reliable within the SCLM.

5.      Power-On Button / Cleaning Cycle Button: Operates as anticipated. Process initiates with the touch of a button.

6.      Temperature Sensor: Since the probe-style sensor is invasive to the brewing process—at least, in the sense that it must be inserted into the brew with each trial—the sensor is at a risk of obstructing the mechanical stirrer’s flange. However, this complication was predicted and does not interfere with the overarching tea-making process.

Comparison of Results to Initial Requirements:

All the requirements and designer expectations were fulfilled with the SCLM prototype’s electronics system. The only concern that may influence changes to the SCLM circuitry is that the current microprocessor—the Arduino Uno—is completely out of input / output pins. Though not currently posing as an issue with the completed prototype, this lack of available ports removes the possibility of integrating further components into the SCLM. As such, further SCLM prototypes should consider a more robust choice of microprocessor.

Tubing System:

Designer Expectations:

The tubing system’s goal is to pass flow from liquid containers and into the carafe through a network of tubing, pumps, valves, and flow meters. With the Two Separate Tubing System (the design that was ultimately selected), fluid travels first from its container, through a solenoid valve, then a pump followed by a flow meter, and finally ending in the carafe. Depending on the specified quantity of servings desired, the valve, pump, and flow meter can all be programmed to accommodate variable volume sizes. The SCLM’s tubing system is graphically depicted in Figure 4.4 below.

Test Conditions and Results:

Testing for the tubing system was carried out in two stages. The first stage occurred early on in the prototype’s development. The following tests and results were performed and obtained:

·         Checking for leaks by passing fluids through an unpowered system

o   This test was performed to assess the risk of potential leakages by examining the means by which tubing subcomponents were attached (i.e. hose barbs and hose clamps).

o   No leakage occurred.

·         Powering solenoid valve on and off to test that flow can be properly started and stopped

o   Tests demonstrated that the solenoid valves work properly.

o   Liquid flow can be arrested as desired.

·         Powered and tested flow meter with pump

o   Flow meter successfully reads liquid volume.

o   Known liquid quantities were dispensed and used to calibrate the flow meter’s precision.

o   According to specification sheet, up to 3% error is expected.

The second phase of piping system tests was conducted using the fully assembled SCLM prototype. The tests performed and the results yielded are as follows:

·         New pumps were added (specifications included in the Appendix) in the hopes of bolstering the fluid flow within the tubing system.

o   New pumps yielded significantly greater flow

·         Complete tea-brewing cycles were run with the new pumps and it was discovered that the SCLM’s solenoid valves were unnecessary. Given the addition of the new pumps, flow is completely arrested if the pumps are not running.

o   Solenoid valves were removed from the piping assembly

·         Piping system is extremely robust and operates as intended throughout every trial.

Comparison of Results to Initial Requirements:

The tubing system meets and exceeds initial requirements and designer expectations. While it was initially believed that the solenoid valve would be necessary to control the flow in the event that the pump is not running, the addition of new pumps removed the necessity of having valves. As such, the resulting final piping system operates on fewer parts than the original without suffering from performance losses.

Liquid Dispenser:

Designer Expectations:

The liquid dispenser—while related to the tubing system—specifically refers to the assembly that consists of the liquid container, the liquid container base, and the quick-disconnect mechanism. Of these three subcomponents, the quick-disconnect mechanism is the most challenging. As discussed in Chapter 3, the quick-disconnect design initially took on the form of the OSVA (a design heavily inspired by water tanks used in automated coffee machines). The function of this quick-disconnect mechanism was to facilitate easy removal of the liquid containers from the SCLM. In this manner, the milk container can be removed following every usage and kept fresh in the refrigerator. Alternatively, the water container can be easily swapped into the milk’s slot and used to flush out the milk piping.

Test Conditions:

The liquid dispenser underwent two phases of testing: one centered on the OSVA and the other on the quick-disconnect valves discussed in Chapter 3.

Phase 1:

After assessing the costs of 3D printing such an apparatus out of non-porous material, a design change overhaul based upon espresso water tanks was implemented and featured:

·         Minimalistic adapter profile and height

·         Usage of O-Rings to ensure proper sealing

·         Integration of the spring and stopper into adapter body

Since it details much of the process in designing the final quick-disconnect solution, testing of the OSVA’s second iteration is well-documented in Chapter 3. As previously explained, an oversight in the second iteration’s sealing design resulted in leakage between the adapter body and the mating sleeve interface. Ultimately, Phase 1 testing could not be fully completed due to time constraints and the OSVA remains incomplete.

Phase 2:

Setbacks encountered during Phase 1 testing created an atmosphere of pressing urgency to locate a viable quick-disconnect solution. Following the recommendation of Dr. James Babcock, off-the-shelf quick-disconnect valves from Colder Products.  After mounting valve inserts onto the liquid dispensing base and outfitting Nalgene containers with valve bodies and pipe-threads, Phase 2 tests were conducted. These tests consisted of:

·         Leakage assessments upon insertion of liquid containers onto their bases

·         Leakage assessments upon removal of liquid containers from their bases

·         Maximum fluid flow assessment

o   Performed by simply attaching the valve insert to the valve body and actuating fluid flow

Results from Phase 2 and Comparison to Initial Requirements:

Following the implementation of the off-the-shelf quick-disconnect valves, leakage between the valve body and valve insert was largely eliminated. No observable leakage resulted upon insertion of the two components. However, a minor amount of leakage occurs upon removing the valve body from the valve insert. Despite this issue however, Phase 2 tests met the initial designer expectations. The following table outlines the maximum flow that can be expected from using the off-the-shelf quick-disconnect valve determined by simply mating the valve body and the valve insert and examining the time it takes for a 1L liquid container to empty.

Table 4.1: Maximum Flow Rate achievable with simple gravity flow and the quick-disconnect valve

As Table 4.1 demonstrates, the off-the-shelf quick-disconnect valve enables a sufficient quantity of flow from the liquid dispensers.

Solids Dispenser:

Designer Expectations:

The solids dispensers’ purpose is to accurately measure and deploy the proper amounts of tea leaves and masala powder based upon the user’s desired serving size. Due to the nature of tea and masala particles, jamming was predicted to be a highly probable issue: one that would increase in severity with the quantity of ingredients contained in the dispensers. On the other hand, the consistency with which dispensing occurs as the volume of ingredients in the container decreases is questionable. After all, decreased volume in the ingredient reservoir directly correlates with the less force being applied to guide ingredients into the slots on the dispensing wheel. In this scenario, incomplete servings would accompany low ingredient levels. The following tests address these potential issues along with accompanying design solutions.

Test Conditions:

Solids Dispensers: 

·         The tea and masala dispensers were evaluated during SCLM’s initial and later development stages. Initial testing was conducted using an external variable power source (ELENCO, MODEL XP-581A) and aimed to pinpoint the cause of jamming. As the prototype began to be consolidated, jamming tests were powered with Pololu geared motors.

·         Upon successfully identifying the source of jamming and implementing a design solution (to be discussed in the following section), the second stage of testing began which involved the completed SCLM prototype and full tea-brewing cycles.

Funnel:

·         The full solids dispensing process is only complete once the solids reach the brew. After the ingredients are dispensed from the solids dispenser, the solids are deployed onto a funnel and travel into the carafe using a fluid-flush assist from the liquids dispensing system.

·         Using the completed SCLM assembly, full tea-making trials were run to determine whether or not all solid ingredients were being properly deployed into the carafe.

Results:

Investigations of the cause of jamming extended into the assembly of the final SCLM prototype. Several insights were made into factors that contributed to jamming along corresponding design solutions were implemented to counter the issues.

·         Since masala particles are exceptionally fine, jamming in the masala dispenser was due to the high-friction interface between the dispensing wheel and the ingredient reservoir.

o   To remedy this problem, the ingredient reservoir’s bottom surface was outfitted with an acrylic layer to provide a low-friction surface upon which the dispensing wheel could spin (see Figure 4.9).

·         Jamming with respect to the tea dispenser is due to tea particulates getting trapped in the space between the dispensing wheel’s outer diameter and the ingredient reservoir’s inner diameter.

o   A funnel attachment was fitted onto the tea dispenser’s ingredient reservoir to ensure that the tea would fall only into the circular slots on the dispensing wheel (see Figure 4.11).

Following implementation of the above design solutions, both the tea and masala dispensers’ performances were vastly improved. However, complications arose even after releasing the proper quantities of solids. Upon reaching the funnel, solids would occasionally cling to the funnel walls—despite the fluid flush assist. Since the SCLM is expected to perform consecutive tea-making trials, every trial succeeding the first iteration will be met with residue liquid on the ingredient funnel from the previous cycle’s fluid flush thereby resulting in inconsistent dispensing. To alleviate this problem, the solids dispensers were outfitted with guide slides to assist the solids in reaching the funnel exit aperture. In this way, the assisting fluid-flush will also have a more focused target to wash into the carafe.

Comparison of Results to Initial Requirements:

Following the testing phase, the solids dispensers operated at full functionality and in complete accordance with the original designer expectations. The completed SCLM prototype is capable of brewing consistently good-tasting Chai Latte only due to the robustness of the entire testing process and the resulting solid dispensers. For future prototype designs, the inclusion of a stepper motor over the current geared motor and end-stop switch combination would be ideal: embodying a design with few components while maintaining full functionality.